See what else came in today at byblos.digital →

Prediction markets went from a rounding error in VC writing to one of the most-discussed themes in the corpus in eight quarters. Attention is at or near an all-time high, capital has rerated the category by roughly 20x in eighteen months, and there is a live, competitive market for early-stage cheques — we count more than thirty distinct companies funded since January 2024, most of them at seed or below.

The thing to know if you are raising this month: the risk narrative has reshaped lately for sure. Federal regulatory acceptance is largely settled. At the same time, state-level litigation, banking access and “social harm” scrutiny are not.

How much are VCs actually talking about this?

Share of voice, normalised per 1,000 articles published, so growth in our corpus doesn’t inflate the trend:

Three distinct phases, and they matter because they explain what any given investor already believes:

1. The election spike (2024)

Zero to 22 per 1,000 in four quarters, peaking alongside the US presidential election. Prediction markets were a novelty story about betting on politics.

2. The trough (H1 2025)

Attention collapsed by two-thirds, to 5.97. The election was over and the category normal again.

3. The institutional re-rating (H2 2025 – now)

From 5.97 to 37.74 in three quarters — a 6x recovery that took the theme well past its election peak. This phase is about CFTC-regulated exchanges, NYSE-parent capital, and sports volume.

For scale, prediction markets now run at roughly the same share of voice as crypto derivatives and open interest (23.72 vs 25.13 in Q2 2026) — a category with a decade of institutional history behind it.

Who is writing about it

Publisher concentration matters more than raw volume when you are building a target list. A fund that devotes a third of its output to your category is a different prospect from one that mentioned it twice.

Castle Island Ventures and Galaxy Digital are the standouts — because prediction markets consume a third to nearly half of everything they write, sustained across 7–13 separate months, and both posted within the last fortnight. That is a durable thesis.

a16z and Insight Partners publish a lot about prediction markets in absolute terms but it is 4–5% of their output. They are interested; not specialists.

Get VC/market research signal every day!

How to get started:

🟣 Sign up and get a free 7-day trial

Get Free VC Intel

🟣 Attend a 1:1 setup call with us
🟣 Build your watchlist of 1,400+ VCs

The deal layer

More than thirty distinct companies with prediction-market rounds since January 2024. The distribution is what should interest you:

Mega-cap — the comparables you will be measured against

• Kalshi — $1B Series F at $22B post-money, led by Coatue, with Sequoia, a16z and Morgan Stanley. 1Valuation path: $2B (Jun 2025) → $5B (Oct 2025) → $22B (May 2026). Roughly 11x in eleven months.

• Polymarket — ICE committed up to $2B at an $8B pre-money, across two tranches. Reportedly seeking a further $1B at north of $20B.

The middle — where a category becomes an industry

• Novig — $75M Series B at $500M, led by Pantera with Multicoin, NFX, Forerunner (sports)

• ProphetX — $35M, regulated infrastructure for enterprise clients

• EDGE Markets — $29.2M Series A, led by CoinFund, Mantis VC, StepStone

• Onyx Odds — $20M Series A at $220M, led by Kraken (sports)

• Opinion — $20M pre-Series A, Hack VC, Jump Crypto, Primitive, Decasonic

Early stage — the live market for first cheques

• The Clearing Company — $15M seed, led by USV, with Haun, Variant, Coinbase Ventures (ex-Polymarket team)

• Limitless — $10M seed, Coinbase Ventures, DCG, 1Confirmation

• River Markets — $8.5M, Haun Ventures, Coinbase Ventures, Y Combinator (prime brokerage)

• Thea — $8M, Maven 11, Spartan, Hack VC

• MNX — $6.4M pre-seed, Manifold Markets co-founders

• Speed Labs — $6.5M seed, Parlay Capital (sports API infrastructure)

• XO Market — $6M seed, Picus Capital and 20VC

• Pascal — $9M, trader-focused exchange

• Adjacent — $2.5M, Night Capital with VanEck, Maven 11, DCG

• PRED — $2.5M seed, led by Accel, with Coinbase Ventures

• Kairos — $2.5M from a16z Fund V (prediction market terminal)

• Gondor — $2.5M pre-seed, Maven 11, Castle Island, Prelude (borrow against Polymarket positions)

• Elastics — $2M pre-seed, ElevenLabs and a16z (AI agents trading prediction markets)

• Giggles — $1.2M pre-seed, led by 1kx

• Pumpcade — $1M pre-seed, led by Pump.fun (markets embedded in livestreams)

• Totalis — Y Combinator’s first all-stablecoin investment

The structural point: the funded set has moved past exchanges into picks-and-shovels. Prime brokerage (River), clearing (The Clearing Company), credit against positions (Gondor), index products (Adjacent), terminals (Kairos, Fireplace), liquidity infrastructure (Fortune Protocol), sports APIs (Speed Labs). That is what a maturing category looks like, and it widens the set of theses a new entrant can credibly occupy.

Investors who have done this more than once

Repeat participation is the strongest available signal that a firm has an internal thesis rather than a oneoff allocation.

Coinbase Ventures is the most active early-stage participant in the category and has never led. Five companies, zero leads. That is a specific, actionable fact: they are a strong participant to build a round around, but you will need a lead elsewhere.

Paradigm and Founders Fund are the opposite — two companies each, but four leads apiece. They lead, and they have concentrated on the two winners rather than spreading across the field.

Haun, Y Combinator and Coinbase Ventures all posted new prediction-market activity within the last ten days.

What has changed in the last four weeks

This is the part that would not show up in a dataset alone, and it is why the framing of your raise matters right now. Since late July:

Set against continued institutional momentum: Cantor began offering brokered access to Kalshi (14 Aug), Robinhood’s CEO called event markets the firm’s fastest-growing unit (29 Jul), Binance.US signalled entry, and ICE said it would look at Polymarket’s new round (19 Aug). Citizens Bank puts the industry at a $3B annual revenue run rate with a path to $10B by 2030.

The two arcs are running simultaneously. Capital is still arriving; the legitimacy fight has moved from Washington to Albany, Olympia and City Hall.

Note also that Kalshi’s and Polymarket’s CEOs jointly backed a $35M fund targeting this sector. The incumbents are funding their own ecosystem, which is an opportunity for infrastructure and a competitive fact for consumer exchanges.

Reading the room

If you are pitching this quarter:

The strongest audience is the specialist publishers — Castle Island Ventures and Galaxy Digital devote 35–43% of their writing to this and are current to within two weeks. They have done the thinking and will not need educating.

The most active early-stage syndicate partner is Coinbase Ventures, five companies deep, but you will need to source a lead separately.

The most credible signal for a non-crypto-native positioning is Accel leading PRED’s seed and Union Square Ventures leading The Clearing Company. Both are generalist firms taking early positions — useful proof that this reads as fintech infrastructure, not only as a crypto trade.

The comparables to price against depend on what you are building. Exchanges are being measured against Kalshi and Polymarket, which is unhelpful at seed. Infrastructure companies — clearing, brokerage, credit, data, terminals — are pricing at $2.5M–$15M seeds with real syndicates, and that is a far more winnable comparison set.

Method: Byblos Digital indexes VC-published content across email newsletters, RSS feeds and firm research blogs — 61,691 articles after cleaning, from 2,217 publishers mapped to 3,739 firms. Share-of-voice figures are normalised per 1,000 articles per quarter against a taxonomy-tagged Prediction Markets theme. Funding events are extracted as structured records and deduplicated by company and round size; counts reflect rounds reported in VC-published content and indicate investor attention rather than a complete market census. Corpus current to 18 August 2026. News developments in section 5 verified against external reporting on 21 August 2026.

Get all Byblos Links

That’s the data. Now go build something.

— Byblos.digital

Subscribe now

See what else came in today at byblos.digital →

Go premium. $10.99/month or $109/year, with a 7-day free trial. You get the full intelligence layer:

• Exclusive research from our database of 1,400 investors and 60,000 newsletters

• Weekly briefs on where capital is actually moving

• Direct 1:1 access to us

Get 7-day free trial

The outlook

This week we read 409 pieces from 207 firms. I guess everyone is coming back from vacations.

Get the last week’s brief to compare!

How to read the documents

The Weekly Brief

What it is. The main narratives investors wrote about this week — each with a plain summary, the firms involved, and the deals behind it.

How we build it. We read every newsletter and blog from the funds we track, group what they wrote into narratives, and tag each as NEW or CONTINUING.

What lifecycle means. A narrative starts when multiple firms write about it, grows into consensus as more publishers pile in, then fades.

Where the links go. Every fund name links to that fund’s card on our site; deal and article links open the source piece on the Byblos feed.

How to use it. See what investors care about right now. Watch NEW tags to catch themes before consensus, and CONTINUING counts to see which last.

The Fund Activity Report

What it is. One card per fund — what they wrote this week, their focus over three months, and how to talk to them.

How we build it. We pull everything a fund published this week against their normal posting rate. Louder means more than usual — a fund getting louder on a theme is paying more attention to it.

How to use it. Find the fund you want to pitch, read their three-month focus and talking points, and open with what they actually wrote. That’s the difference between a cold pitch and a relevant one.

Same deal with the hyperlinks on fund names leading to their VC cards on our app.

Want your VC/market research automated?

How to get started:

🟣 Sign up over here and get a free 7-day trial of all features

Get free Byblos

🟣 Attend a 1:1 setup call with us
🟣 Receive free exclusive briefs from the 1,400+ VC database and use the app freely

If you don’t have time

Here’s the short version.

Read more

See what else came in today at byblos.digital →

Prefase. Last week I had a call with a founder who turned out to be completely fine off my services. She didn’t need my investor research and pitch guidance. Her VC list was rich, she had 5 calls/week, tailored pitches, and a good deck.

It pushed me to think how I can help her further in guidance and advising along the way.

This time I searched 12,593 fundraising and founder-advice articles published by VCs since January 2024 and counted which complaints come up again and again. This isn’t what I think founders get wrong but what investors keep writing about, in order of how often.

One thing stands out before the list. Almost all of these are process mistakes, so nobody is losing a round because their market is too small (yay!). It happens much more often when they contact 10 investors, send a deck nobody could skim, and then go quiet for six weeks (that’s sad gang).

1. Generic outreach to too few people

233 articles. The most common complaint by a wide margin.

Sending the same message to every fund with the name swapped in. Investors spot it immediately, and it tells them you haven’t read anything they’ve written.

The bigger version of the problem is contacting too few investors. Hustle Fund wrote about this in August. Their guidance is a list of 150 to 300 relevant investors, built before you send a single email. “Relevant” is doing the work in that sentence. A list of brand names who don’t fund your stage isn’t a list.

The fix is volume and specificity together. Reference something the fund actually published, and be right about why they’d care.

2. Metrics that don’t hold up

191 articles

Picking the one chart.

Investors have seen that shape before and will ask what happened to the others. Registered users, impressions and waitlist size presented as traction, with no retention or revenue number behind them. The missing reasoning is the problem.

If a number is weak, say so before they find it. Volunteering a bad number buys you more than a good one you had to be cornered into.

3. The deck as a data dump

176 articles

Slides packed with metrics, acronyms and half-finished TAM maths, hoping the numbers speak for themselves. Jacob Mullins wrote in July that without a story tying it together, the data feels hollow.

Airbnb’s Sequoia pitch is the example investors keep citing. Hotels were failing travellers, people wanted something more authentic, and the founders couldn’t make rent. The numbers supported the story rather than replacing it.

A quick test: if someone reads only your slide headlines, do they get the argument? If not, you’ve written a document, not a pitch.

Get VC/market research signal every day!

How to get started:

🟣 Sign up and get a free 7-day trial

Get Free VC Intel

🟣 Attend a 1:1 setup call with us
🟣 Build your watchlist of 1,400+ VCs

4. Not knowing your own numbers

141 articles

Getting defensive or vague under a sharp question. I will stand for this forever because investors invest in people, not in projects. They will read your reaction, read your expressions, read your attitude. Over-preparing the deck and under-preparing yourself will become a disadvantage. A good deck won’t survive a question you can’t answer.

The Collison brothers come up repeatedly here. Not flashy presenters, but they knew the business well enough that every question made investors more confident rather than less.

Rehearse the questions, not the script.

5. Going quiet

116 articles

No follow-up after a meeting that went well. Investor silence is often a test of persistence rather than a rejection: what are you going to do? Same with not sending updates to investors who passed. I did see a recent X debate on this topic but funds pass on timing all the time, and a monthly update is the cheapest way to turn a “not now” into a “now.”

Don’t only get in touch when you need something. Investors notice founders who appear exclusively when raising.

6. Inflated or top-down TAM

112 articles

“It’s a $50B market and we only need 1%.” This is the line VCs write about most often as a warning sign. Top-down sizing pulled from an analyst PDF will just tell them you haven’t done the work. Bottom-up — customers times price times frequency — tells them you have. A smaller number you can defend beats a large one nobody believes.

7. Talking valuation before conviction

102 articles

Leading with terms in a first conversation, before anyone believes in the company yet. Anchoring to a peer company’s headline valuation without having their metrics.

Pricing a round so high that the next one has to be a down round. Try thinking two rounds ahead like an investor would.

8. Raising too late

51 articles

Starting a raise with a few months of runway left. It’s a weak negotiating position basically. The common guidance is to start 9 to 12 months out and treat it as a pipeline with a lead time, not an event.

Running out of cash mid-process means taking the first offer instead of the right one.

Even in my investor prospecting work, the extra time we take is 2-4 weeks to process the VC research work and settle the deal with a client. If you plan a raise, plan it well.

9. Pitching at investors instead of thinking like them

Recurring theme across the pitch-advice pieces, not counted separately

Treating a VC like a professor marking a paper. This isn’t really a scoring system but them working out whether they’d regret passing.

Their actual questions are simpler than founders expect. Will this person keep going when it gets hard? Will this company attract talent and customers? Will this return the fund?

Founders who miss this pitch features and valuations instead of inevitability.

10. Mistaking enthusiasm for commitment

From a16z’s GTM writing, not counted separately

Find out early who decides and what the process is. Ask directly. Who else needs to weigh in, and what happens between now and a decision.

The fastest way to stall a raise is to spend six weeks with a fan instead of a decision-maker.

The pattern

Nine of these ten are research and process problems. That’s useful, because they’re all fixable before you send the first email.

The first one matters most. Every other mistake here gets worse when you’re pitching an investor who was never going to fund your stage, your sector or your geography.

This is what I strive to help founders with, researching which funds will actually want to hear their pitch based on their writing and moves.

Method: 12,593 articles tagged as venture capital, fundraising or founder advice, published by VC firms and investors between January 2024 and August 2026, from the Byblos corpus of 61,691 VC-published pieces across 2,217 publishers. Categories were counted by pattern-matching article text. Frequencies show how often each theme is discussed, not how serious it is. Sources include a16z speedrun, Hustle Fund, Techstars, GTMfund and Jacob Mullins.

Get all Byblos Links

That’s the data. Now go build something.

— Byblos.digital

See what else came in today at byblos.digital →

“AI” is not one market. The category you put yourself in changes how much you can raise. Move one category to the left or right, and the median round size changes by about 5x. The number of investors willing to write a first cheque changes by even more.

I read 61,691 pieces of content that VC firms published between January 2024 and July 2026. That means newsletters, RSS feeds, and firm research blogs. From that, I pulled 239 funding rounds across five related AI verticals. Here is what the money did.

Two numbers do most of the work here. AI video has the highest median round of the five, at $42.7M. Only two of its ten priced rounds came in at or under $10M. Agentic payments has the lowest median, at $8.3M. Six of its seven priced rounds were $10M or under.

It is the same technology wave. The funding behaviour is completely different.

One note on the table. Some rounds did not disclose a size. The "≤ $10M" counts use only the rounds that did.

Get VC/market research signal every day!

How to get started:

🟣 Sign up and get a free 7-day trial

Get Free VC Intel

🟣 Attend a 1:1 setup call with us
🟣 Build your watchlist of 1,400+ VCs

Why AI video raises big rounds

The vertical splits into two groups. A few very large rounds, and a thin set of small ones. The large ones are Synthesia's $200M Series E at a $4B valuation, Higgsfield's $66M, Tavus's $40M Series B, and Hedra's $32M Series A led by a16z's Infrastructure Fund.

Here is why.

Read more

See what else came in today at byblos.digital →

Among the founder calls and LinkedIn outreach, I decided to go and use my own Byblos App. Instead of reading thousands of VC newsletters, I skimmed through the intel section. The part that caught my attention was a piece by Edge & Node — a crypto fund, not even an AI one — wrote that good-enough AI is getting cheap faster than the data centers can pay for themselves.

That didn’t line up with anything else I’d read because everyone else was writing about the huge amounts of money going into those data centers.

So I went back and checked everything.

Everyone writes about money (surprise)

Nvidia announced a $500 billion financing program for AI compute, with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR.

Four separate funds flagged this too— Nazaré, TEN31, Haystack and Kucoin.

Among the deals, CoreWeave is leasing A100 GPUs through 2029. Riot Platforms is building a $9.1 billion data center with Anthropic as the tenant. Hut 8 signed a fifteen-year data center agreement in Texas.

Fifteen years.

That’s an office building contract. You sign that when you’re confident the thing you’re renting out is worth roughly the same in year twelve as in year one.

Who writes about prices?

I found basically one person writing about what AI actually costs now.

Chris Zeoli:

• AI token prices falling around 10x per year;

• Alphabet cut the cost of serving Gemini by 78% in twelve months;

• GPT-3-class pricing down 1000x;

• DeepSeek published a 545% cost-profit ratio on serving.

Kucoin picked up the same shift in a different way as they wrote that the market has stopped asking how much companies spend on AI and started asking how efficiently they spend it.

That’s it. Ten funds on the financing. Basically one guy on the deflation.

I think that point is that these two things haven’t been put next to each other yet.

Get VC/market research signal every day!

How to get started:

🟣 Sign up over here and get a free 7-day trial of all features

Get free Byblos early

🟣 Attend a 1:1 setup call with us
🟣 Build your own watchlist from 1,400+ VC database

Proof-reading my content

As I said before, bitcoin miners are the ones actually building these data centers.

1. Bitfarms renamed itself Keel Infrastructure and quit US mining entirely.

2. Bitdeer converted a mining site for a company called Volta.

3. Galaxy bought three data center sites in Texas.

So the people who last built giant power-hungry facilities for an asset that crashed are building giant power-hungry facilities again. Except this time they’re landlords: they collect rent and someone else takes the price risk.

I think this is not just luck but that they’ve done this before and learned something.

Bubble (?)

I want to be fair, because the bull case is sitting in the same data.

Zeoli’s other number: Google’s token consumption grew 330x in two years, so price down 10x, volume up 330x. This basically means the revenue still goes way up.

But volume growth goes to whoever owns the customers. Price deflation hits whoever sells the compute. That’s two different groups.

Almost nobody is bearish

I went looking for the bear case too, and found Michael Burry — short Micron, short Nebius, short Oracle, short the semiconductor ETF, warning that AI infrastructure demand isn’t sustainable.

I’m not saying Burry’s right. I’m saying if you only read VC newsletters, you would think there is no disagreement here at all if four funds on the financing, one on the short.

Why I am writing this

Figure out which side of the price curve you’re on.

If you resell compute, or your product is a thin layer on someone else’s model, the price you can charge falls as the model gets cheaper. You don’t capture the volume growth either, because the customer isn’t really yours.

If you own the customer, deflation helps you. Costs drop, you keep the margin.

And if you’re building physical infrastructure, Kucoin’s point is the one I’d underline. They wrote that power scarcity is now the main value driver for neoclouds. Chips get cheaper every year but a grid connection doesn’t.

One last thing I didn’t expect to find. Kucoin tracked DRAM prices surging, Apple testing Chinese CXMT memory for iPhones and MacBooks, and memory becoming a serious chunk of the iPhone 18 Pro’s bill of materials. The AI buildout has made ordinary memory expensive.

So if you build hardware, your costs went up last quarter because of AI. Even if your product has nothing to do with AI.

Get all Byblos links

That’s the data. Now go build something.

— Byblos.digital

Subscribe now

This year bitcoin is down, and Google searches are down compared to last year. And yet the team is already planning 2027 and 2028 events.

I sat down and talked to Samuli Rautiainen, their CMO, about how they built it.

Nobody had done it, and they had the right people to try

Before BTCHEL there was no large-scale Bitcoin conference in the Nordics. People working on Bitcoin in the region simply had nowhere to meet each other in person, which is a strange gap for an industry that runs on people knowing each other.

So the team that ultimately became BTCHEL reasonably asked themselves whether they should make one.

The idea came from experience. Remu Karhulahti co-founded the Bitcoin Association and had studied Bitcoin for years, so he came in with a network already there. Luke de Wolf brought deep credibility in the space – a cybersecurity expert who’s consulted for major clients in oil & gas, manufacturing, maritime, and logistics in Finland, Norway, and Canada.

When I asked Samuli why their event went better than comparable ones (e.g., the Bitcoin symposium in Sweden), he gave four reasons:

• Timing

• A team with real depth and an international network

• A strong event producer

• Young people running it

I guess the third one is the most boring but it totally makes sense. Knowing Bitcoin and knowing how to run a room of a thousand people are two different skills. You can be an expert in the field but be terrible as a founder. That wasn’t the case here.

Want your VC/market research automated?

How to get started:

🟣 Sign up over here and get a free 7-day trial of all features

Get free Byblos early

🟣 Attend a 1:1 setup call with us
🟣 Receive free exclusive briefs from the 1,400+ VC database and use the app freely

They sold it before they built it

I want to highlight the order they did things in.

Before building anything, they had to find the funds and users. They did both very plainly.

1. The team went to companies and asked them to sponsor the conference. Commitments came in fairly quickly — Coinmotion, Finland’s biggest crypto exchange, was one of the early ones.

2. The team put up a landing page and sent it round their network. 400 sign-ups came in over two months.

Only after both of those did they start building the conference properly.

The reason they could afford to work that way was because nobody was getting paid. The whole team worked for equity at the start, and when I asked Samuli what the hardest part was, that’s what he named — asking people to put months into something on a promise.

Even with sponsors committed, the cash timing on events is awkward. You commit to costs early, because venues and production and speakers all need booking well ahead of time. Attendees buy tickets at the last minute. So you spend real money for months against results you can’t yet see. They broke even in year one anyway.

I liked Samuli’s point on how to make an event a success: You don’t really need connections, you just talk to people and see what happens.

While you’d think a bear market and lack of interest in bitcoin would be a bad thing, the BTCHEL team’s read is that downturns are good ops. It’s part of the cycle, so you work through it rather than pausing and waiting for sentiment to come back. The main proof of this is how the team is already planning the 2027 and 2028 conferences.

What is on Agenda

BTCHEL writes as well — podcasts, newsletters, articles. I asked what three topics are talked about in the community or are worth it from their point of view.

1. Self-custody and security

In late July and early August 2026, a firmware bug in Coldcard hardware wallets let attackers guess recovery seed phrases, and more than $100 million in bitcoin was taken across thousands of addresses. The bug went back to a firmware release from March 2021 — affected devices skipped the proper hardware random number generation, so the seed phrases they produced could be brute-forced. The special reason this became a big story is because Coldcard was the wallet bitcoin maximalists recommended above all others for cold storage – “the safest option failed”.

What makes it worth writing about rather than just reporting is the fix, which isn’t obvious. Updating the firmware doesn’t help a seed that was already generated badly. If you were affected you need to set up a completely new wallet with a fresh seed and move your funds across; moving the old backup onto a different device carries the same entropy problem with it. There’s also been a wave of phishing attacks aimed at people in the middle of migrating, so links are worth treating carefully. It’s the kind of fast-moving, high-stakes story that shows exactly why timely, trustworthy Bitcoin coverage matters.

2. Developer split

There are proposals on the table for changes to the Bitcoin code and genuine disagreement about them, alongside a debate about whether AI investment is feeding negative sentiment around Bitcoin.

3. Bitcoin loans.

More holders don’t want to sell. Samuli’s position is that one of two things has to give. Either governments stop taxing the sale, or borrowing against bitcoin becomes a normal option: You back the loan with your bitcoin and pay interest instead of selling it.

That last theme runs straight into what’s happening at the event. The category with the most companies behind it is custody and safe storage, which makes sense both as a business case and for self-sovereignty, which is about not having to trust a bank with your money. The expo isn’t early-stage companies pitching, and the speakers are on live industry issues rather than fundamentals.

What stuck with me is the way the BTCHEL team proved it would work before building it. They didn’t plan the big launch based on guesses but literally put up a landing page, sent it to their network, and acted based on sign-ups and conviction.

They treated demand as the first experiment, not the last. For anyone building something new, that’s the lesson: Your first job isn’t to build the most brilliant product but to find the cheapest possible feedback that proves someone wants it.

Let me know what you think of this format in the comments! While I found it useful for myself, I am very much interested in what my audience thinks. Drop the people I should reach out for interviews or the topics you’d like to read on!

That’s the data. Now go build something.

— Byblos.digital

Subscribe now

See what else came in today at byblos.digital →

Go premium. $10.99/month or $109/year, with a 7-day free trial. You get the full intelligence layer:

• Exclusive research from our database of 1,400 investors and 60,000 newsletters

• Weekly briefs on where capital is actually moving

• Direct 1:1 access to us

Get 7-day free trial

The outlook

This week we read 246 pieces from 135 firms. Here is what investors wrote about, the deals behind it, and who to talk to.

Get the last week’s brief to compare!

How to read the documents

The Weekly Brief

What it is. The main narratives investors wrote about this week — each with a plain summary, the firms involved, and the deals behind it.

How we build it. We read every newsletter and blog from the funds we track, group what they wrote into narratives, and tag each as NEW or CONTINUING.

What lifecycle means. A narrative starts when multiple firms write about it, grows into consensus as more publishers pile in, then fades.

Where the links go. Every fund name links to that fund’s card on our site; deal and article links open the source piece on the Byblos feed.

How to use it. See what investors care about right now. Watch NEW tags to catch themes before consensus, and CONTINUING counts to see which last.

The Fund Activity Report

What it is. One card per fund — what they wrote this week, their focus over three months, and how to talk to them.

How we build it. We pull everything a fund published this week against their normal posting rate. Louder means more than usual — a fund getting louder on a theme is paying more attention to it.

How to use it. Find the fund you want to pitch, read their three-month focus and talking points, and open with what they actually wrote. That’s the difference between a cold pitch and a relevant one.

Same deal with the hyperlinks on fund names leading to their VC cards on our app.

Want your VC/market research automated?

How to get started:

🟣 Sign up over here and get a free 7-day trial of all features

Get free Byblos early

🟣 Attend a 1:1 setup call with us
🟣 Receive free exclusive briefs from the 1,400+ VC database and use the app freely

If you don’t have time

Here’s the short version.

One fund noticed the AI capex problem. Seven noticed the wallet exploit. Nearly every capex data point in the brief was flagged by a single fund — Activant on inference economics, Hamilton Lane on the Nvidia backstop, Chris Zeoli on memory. The Coldcard bug was independently flagged by Galaxy, Bankless, Edge & Node, Apollo, Pompliano, Lark Davis and Rob Behnke. Conclusion: the AI capex narrative is a handful of analysts talking; crypto security is genuine consensus. If you’re raising on AI infrastructure you’re pitching individuals. On security tooling you’re pitching a room that already agrees.

Read more

See what else came in today at byblos.digital →

I read what 500+ VC firms published this week. This time, I was interested in what funds stopped talking out.

Three things stood out: DeFi, Solana, Climate. I checked whether the funding data matched, and it did.

Get VC/market research signal every day!

How to get started:

🟣 Sign up over here and get a free 7-day trial of all features

Get free Byblos early

🟣 Attend a 1:1 setup call with us
🟣 Receive free exclusive briefs from the 1,400+ VC database and use the app freely

DeFi

DeFi funding has fallen for three straight quarters. Q2 2026 was the weakest quarter since late 2023. The number of DeFi deals hit its lowest point since 2020.

The reason is not that DeFi failed. Lending, exchanges, and stablecoins all function. Investors stopped writing about it because it already works. They write about problems that are not solved yet.

The type of DeFi that gets funded also changed. In 2021, money went into protocol governance tokens. In 2026, money goes into companies with revenue that follow financial regulations.

If you build in DeFi, do not describe your company as a “DeFi protocol.” Describe it as financial infrastructure with revenue.

Climate

Climate tech mentions dropped, but climate tech funding went up. It reached $26.1 billion in the first half of 2026, up 55% from the year before.

It looks like the money stopped going to projects labeled “climate.” It went to data centers. Low-carbon data centers went from 3% of climate funding to 34% in one year. The data center category grew more than 800% and became larger than energy. Funding for carbon projects fell 61%, the lowest since 2020.

The grid, the power, the cooling, and the batteries did not lose funding, but now they are called “AI infrastructure.”

If you build energy or power products, describe it as AI infrastructure.

The pattern

So who is actually deploying right now? A few names, from 2026’s funding data. Coinbase Ventures has been the most active crypto investor this year — 12 deals in Q1 alone, more than double the next firm. Paradigm raised a new $1.2 billion fund, which means they have to deploy it. a16z, Castle Island, and Galaxy Digital each did around five deals. And traditional firms — Sequoia, Founders Fund, Bain — showed up in crypto rounds too, treating the downturn as a buying opportunity.

Get all Byblos Links

That’s the data. Now go build something.

— Byblos.digital

Subscribe now

See what else came in today at byblos.digital →

The consensus

Ask what’s happening in Bitcoin right now and you’ll get a price. Down about 25% this year, from roughly $87,500 in January to the mid-$60,000s. Spot ETF assets have fallen to around $81 billion, down about 30% since January. Strategy’s enterprise mNAV closed below 1.0 for the first time on 26 June, at 0.99 — the market now values the largest corporate holder at less than the bitcoin it holds.

This story is accurate, and it is the least interesting thing in the data.

Get VC/market research signal every day!

How to get started:

🟣 Sign up over here and get a free 7-day trial of all features

Get free Byblos early

🟣 Attend a 1:1 setup call with us
🟣 Receive free exclusive briefs from the 1,400+ VC database and use the app freely

The turn

Over the same period a basket of Bitcoin mining stocks rose more than 50%. TeraWulf gained over 70%. The companies built to produce bitcoin outperformed bitcoin by roughly seventy percentage points.

Capital did not leave the Bitcoin economy this year. It moved from holding the asset to powering the infrastructure around it — and our corpus shows the rotation happening in three independent places at once.

First, in what VCs write about. Across the email and RSS substrate of our corpus, Bitcoin Treasury Companies fell from 86.3 mentions per 1,000 articles in August 2025 to 29.5 by July 2026 — down 66%. AI Capex went from 64.0 to 149.4 over the same window, up 2.3×. Bullish sentiment on treasury companies dropped from 50.6% to 30.3% while bearish quadrupled from 3.1% to 13.6%, and that step-change has now held for four consecutive quarters.

Second, in what they say is causing it. Three unconnected publications named the same mechanism within two weeks of each other. Apollo Crypto attributed treasury stress to capital rotation into AI trades. CoinShares described investor rotation away from BTC-holding companies toward AI and mining. Jordi Visser, via Gate Ventures, put it plainly: capital is rotating to AI stocks.

Third, and most concretely, in the deals. Deduplicating every miner-to-hyperscaler transaction in our corpus produces roughly $38.8 billion in signed agreements — IREN × Microsoft at $9.7B, CoreWeave’s $9B acquisition of Core Scientific, Hut 8 × Google, AWS × Cipher Mining at $5.5B, Galaxy Digital’s 800MW CoreWeave lease at $4.5B. Bernstein, working from public filings rather than newsletters, puts the sector-wide figure near $150 billion across 7.5+ gigawatts.

A quick note on method, because it matters. Raw mention counts are close to worthless here — a single funding round appears in up to fourteen newsletters, and Bitcoin itself functions as a tablestakes asset the way gold does. Nobody writes “we bought gold.” They write about the mine, the refinery, the ETF, the jurisdiction. Counting the word “bitcoin” measures the weather. Counting deals, instruments, and dollars measures the climate. Everything below is the second kind.

The deep dive: where the money pools

Miners have never raised venture capital.

We pulled every miner financing event in our corpus back to 2022. Riot, Marathon, Bitfarms, CleanSpark, HIVE, IREN, TeraWulf, Core Scientific — fifteen financings, roughly $6 billion, and not a single venture round. Convertible notes, senior notes, share offerings, bitcoin-backed credit lines.

This is worth stating clearly because it is easy to misread: the absence of VC in Bitcoin mining is a structural default. It’s not like a unique 2026 sign.

The insight is what happens when a miner stops mining.

CoreWeave started as an Ethereum mining operation. Crusoe started as flared-gas Bitcoin mining. Both repositioned as AI compute companies, and the capital markets available to them changed completely. CoreWeave: $221M, then $371M, then $3.5B led by NVIDIA and Blackstone, then a $7B Series C. Crusoe: $102M seed, $600M Series D, and a $1.375B Series E in January led by Mubadala alongside Founders Fund and Bain Capital Ventures.

Identical assets. Identical engineering. A different category label, and a different asset class of investor.

And some of this pivot is being funded by selling bitcoin at a loss.

Public miners sold a record 32,000+ BTC in the first quarter of 2026. RootstockLabs’ Sam Golden documented the economics: weighted-average cash cost to mine near $80,000, hashprice compressed to $28–30 per PH/s/day, market price around $64,000. A subset of that selling went specifically to fund AI compute buildouts. Golden’s verdict on the trade — “one of the most expensive ways to raise a dollar” — is the most honest sentence written about this transition.

Core Scientific liquidated 1,992 BTC, about $175M, in March to fund its own.

So the rotation is not metaphorical. Bitcoin is being converted into AI data centres, coin by coin, frequently below cost.

The treasury companies are the other side of the same trade.

NYDIG’s Greg Cipolaro frames Strategy precisely: 847,363 BTC worth $52.4B plus $2.55B cash against $22.2B of debt and preferred claims — roughly $32.7B of surplus NAV. Solvency is not the issue. The issue is that the flywheel requires MSTR above 1.0× mNAV and STRC near $100 to function; STRC has been trading around $90 after touching $74. His conclusion: “a market-access problem, not a solvency problem”.

When that mechanism stalls, the behaviour inverts. Strategy authorised up to $1.25B in bitcoin sales to fund dividends and buybacks. Empery Digital sold half its treasury for $87.1M. BitMine pivoted to ETH staking, which now generates $45.7M quarterly — 98% of its revenue. Metaplanet bought a securities firm. Roughly 1.16 million BTC sits on corporate balance sheets earning nothing, and Cosimo Digital’s Rob Frasca argues the persistent discount isn’t sentiment at all: “It is a structural flaw.” Without a redemption path, no arbitrage can close the gap.

The payoff for founders

Go premium. $10.99/month or $109/year, with a 7-day free trial. You get the full intelligence layer:

• Exclusive research from our database of 1,400 investors and 60,000 newsletters

• Weekly briefs on where capital is actually moving

• Direct 1:1 access to us

Get 7-day free trial

Read more

See what else came in today at byblos.digital →

Go premium. $10.99/month or $109/year, with a 7-day free trial. You get the full intelligence layer:

• Exclusive research from our database of 1,400 investors and 60,000 newsletters

• Weekly briefs on where capital is actually moving

• Direct 1:1 access to us

Get 7-day free trial

The outlook

This week we read 268 pieces from 149 firms. Here is what investors wrote about, the deals behind it, and who to talk to.

How to read the documents

The Weekly Brief

What it is. The main narratives investors wrote about this week — each with a plain summary, the firms involved, and the deals behind it.

How we build it. We read every newsletter and blog from the funds we track, group what they wrote into narratives, and tag each as NEW or CONTINUING.

What lifecycle means. A narrative starts when multiple firms write about it, grows into consensus as more publishers pile in, then fades.

Where the links go. Every fund name links to that fund’s card on our site; deal and article links open the source piece on the Byblos feed.

How to use it. See what investors care about right now. Watch NEW tags to catch themes before consensus, and CONTINUING counts to see which last.

The Fund Activity Report

What it is. One card per fund — what they wrote this week, their focus over three months, and how to talk to them.

How we build it. We pull everything a fund published this week against their normal posting rate. Louder means more than usual — a fund getting louder on a theme is paying more attention to it.

How to use it. Find the fund you want to pitch, read their three-month focus and talking points, and open with what they actually wrote. That’s the difference between a cold pitch and a relevant one.

Same deal with the hyperlinks on fund names leading to their VC cards on our app.

Want your VC/market research automated?

How to get started:

🟣 Sign up over here and get a free 7-day trial of all features

Get free Byblos early

🟣 Get added to our Telegram group
🟣 Attend a 1:1 setup call with us
🟣 Receive free exclusive briefs from the 1,400+ VC database and use the app freely

If you don’t have time

Here’s the short version.

The big picture: two economies ran the week.

• AI infrastructure — huge, but wobbling. Nvidia floated a $250 billion backstop for OpenAI’s data centers and put $5 billion into Safe Superintelligence. Yet the same week, Korea’s market crashed 40%, memory stocks turned bearish, and a star AI fund gave back most of a 439% run. The spending is still enormous; the confidence behind it is cracking. Build faster vs. this can’t last is the tension under almost every AI story right now.

Read more

See what else came in today at byblos.digital →

Go premium. $10.99/month or $109/year, with a 7-day free trial. You get the full intelligence layer:

• Exclusive research from our database of 1,400 investors and 60,000 newsletters

• Weekly briefs on where capital is actually moving

• Direct 1:1 access to us

Get 7-day free trial

I’ve been thinking about how the most important shifts rarely arrive as headlines. The venture firms we tracked this week stopped agreeing on whether AI spending is something to celebrate or something to survive — and that disagreement matters more than any single number in it.

Let me start with the event that made it concrete. Leopold Aschenbrenner — the author of “Situational Awareness,” the essay a good chunk of tech passed around last year — ran an AI-focused fund that reached $45 billion in assets. This week it collapsed under margin pressure and handed roughly $16 billion of its public book to Ken Griffin’s Citadel in a forced sale. I find these moments clarifying. Not because one fund failing predicts anything on its own, but because of how differently people wrote about it afterward.

Quick note. We launched on Product Hunt last week, and came 28th at 43 followers and 22 upvotes. It was a great experience for our team, and, while we believe we can do much more, I wanted to express gratitude for your support. Byblos App exists to take VC and market research off the plate for people who want to build more and better.

Tired of spending hours on investor research?

🟣 Sign up over here and get a free 7-day trial of all features

Get free Byblos App

🟣 Get added to our Telegram group
🟣 Attend a 1:1 setup call with us
🟣 Receive free exclusive briefs from the 1,400+ VC database and use the app freely

The believers are still building the case for scale

One group of investors spent the week reinforcing the idea that in AI, you win by spending. OpenAI moved to secure cloud spending toward $750 billion by 2030. AMD’s stock rose 10% on a new supply deal with Anthropic, which tells you the compute race is widening beyond a single chipmaker. And in the 20VC newsletter, Menlo Ventures described backing Anthropic after stepping outside its own fund rules — the benchmarks were strong enough that the burn didn’t matter. I understand the logic. If you believe the frontier is a land grab, then the compute bill isn’t a cost, it’s the entry fee. Founders pitching this group should lean all the way into ambition. These are people who want to hear that you intend to be the biggest.

The skeptics started asking for the math

At the same time, a different group turned quietly critical. The cost to insure NVIDIA’s debt hit a record 82 basis points as the scale of AI commitments raised real bubble questions. Galaxy Digital titled its research brief “AI’s civil war.” Amadeus Capital titled its newsletter “AI has left the chat” and moved its attention to the physical output of AI — drug discovery, materials science. Kerman Kohli asked, plainly, whether demand for compute is finite, noting that AI and semiconductor names had fallen 30 to 40% from their highs.

What strikes me is that none of these people are AI pessimists. They’re asking the oldest question in investing — does the money come back — and they’re asking it out loud for the first time in a while. A founder pitching this group with a story about out-spending everyone will lose them on the first slide. They want to hear about revenue, and about why your product works without a fortune in compute.

What we think this actually means

Here’s the implication we keep landing on, and it’s the part nobody is stating directly. The AI-spending debate isn’t really about AI. It’s about which founders still get funded on faith, and which now get interrogated on the numbers. And that line moved this week — not in an announcement, but in the accumulated tone of what these firms chose to write. The same company, with the same deck, is fundable by a smaller portion of the room than it was three months ago. If you’re mid-raise, you may have felt a meeting go colder than expected and not known why. This is part of why.

There’s a third group forming, too, for founders who fit neither camp. Mercury Fund wrote “Open Weights and the Era of Vertical AI.” Emergence Capital wrote “Open Weights Are a Bet on Autonomy,” describing a portfolio built on open models paired with proprietary data. Their wager is that the value settles not in raw scale but in open models applied to specific industries. If that’s your company, that’s your room, and it’s a room that barely existed as a distinct thesis a year ago.

The practical version

So what do you do with this. Before you pitch a fund, read what it published in the last month. It will tell you which of these three conversations that firm is having with itself right now. Bring a scale-and-ambition story to a believer like Menlo. Bring a revenue-and-efficiency story to a skeptic like Amadeus or Galaxy. Bring an open-model story to Mercury or Emergence, ideally in their own language. The pitch that wins one of these rooms will actively lose another.

Everything in today’s digest came from the Byblos live feed. The full newsletter archive, the VC directory, and tomorrow’s signal are all waiting for you at Byblos.digital - the investor interned, curated.

Get all the Byblos Links

That’s the data. Now go build something.

— Byblos.digital

Subscribe now

See what else came in today at byblos.digital →

Go premium. $10.99/month or $109/year, with a 7-day free trial. You get the full intelligence layer:

• Exclusive research from our database of 1,400 investors and 60,000 newsletters

• Weekly briefs on where capital is actually moving

• Direct 1:1 access to us

Get 7-day free trial

The outlook

This week we read 297 pieces from 132 firms. Here is what investors wrote about, the deals behind it, and who to talk to.

How to read the documents

The Weekly Brief

What it is. It covers the main narratives investors wrote about this week. Each story has a plain summary, the firms involved, and the deals that back it.

How we build it. We read every newsletter and blog from the funds we track for the week. We group what they wrote into narratives. We tag each narrative as NEW or CONTINUING.

What lifecycle means. A narrative starts when multiple firms write about it. As the number of publishers increases, it grows and becomes consensus, to then fade away.

How to use it. Read it to see what investors are paying attention to right now. Watch the NEW tags to catch themes before they are consensus. Watch the CONTINUING counts to see which themes last and which ones fade.

The Fund Activity Report

What it is. One card per fund. Each card shows what a fund wrote this week, what they wrote over the last three months, and how to talk to them. Look up the funds you care about.

How we build it. For each fund, we pull everything they published this week. Baseline is a fund’s normal posting rate. Louder means they posted more than that this week. A fund getting louder on a theme is a fund paying more attention to it.

How to use it. Find the fund you want to pitch. Read their three-month focus and their talking points. Open your outreach with what they actually wrote. This is the difference between a cold pitch and a relevant one.

We launch our app on Product Hunt on the 30th of July.

Follow our Launch

It would mean a lot if you showed up and interacted with it on the day — a comment or an upvote goes a long way for a small team.

What you can do:
→ Follow every VC in your sector in one subscription
→ Filter your feed by fund, sector, theme, and sentiment
→ Open a VC card for any fund to see their investments, focus, and links
→ Build a watchlist around your raise and track what your target investors are writing

Want to make the batch early? How to get started:

🟣 Sign up over here and get a free 7-day trial of all features

Get free Byblos early

🟣 Get added to our Telegram group
🟣 Attend a 1:1 setup call with us
🟣 Receive free exclusive briefs from the 1,400+ VC database and use the app freely

If you don’t have time

Here’s the short version of what the funds wrote this week. The full brief and fund report are worth it — but if you’re about to walk into a meeting, read this first.

What VCs wrote about most

• The AI IPO race (continuing). The big labs are heading for public markets. Anthropic is valued at $965 billion and started scheduling investor meetings ahead of a possible October IPO, trying to beat OpenAI to the market. OpenAI has confidentially filed and could seek up to $1 trillion. SpaceX already went public in June.

• Open models and distribution (continuing). Chinese open-weight models have gone from novelty to core infrastructure in the US AI stack. Hugging Face is now the main place those models get distributed.

• The agent era (new). AI agents are moving from helping with workflows to running them. The debate among funds: do autonomous agents kill the software layer or reinforce it?

Read more

As we launch Byblos Digital app on the 30th of July (two days away!), I wanted to hop on here to speak about the problems we want to solve, and why you should use our app.

Founder’s job is to build. Raising is hard enough without spending hours on VC research by hand. It has a hidden time-sink. Before a founder ever pitches, they have to figure out who to pitch — which funds are active, which invest at their stage, which care about their sector, and what those investors currently believe. That research is slow, manual, and it's not building the company. Founders do it anyway because getting it wrong wastes the raise.

Here are the complaints we hear most, what you want instead, and how Byblos helps. There is a demo video near the end.

The problems

Tell us how many of these you have hit.

1. You only reach out to Tier-1 investors with your first pitch because they are easy to find. Surely, they almost never respond.

2. You are about to pitch a fund, but the only thing you have read is some blog post definitely not enough to speak their language.

3. You send 100 cold emails and hope that something sticks. You do not know which 10 investors are actually active, and how to approach them.

4. The partner references their thesis you’ve never head of while in the meeting. The only thing you can do is nod.

5. You pitch a fund one angle. Later you find out they moved into your sector last month and wanted a different one.

6. You work off a “top VC” list. Half of it is two years old. Some of those funds stopped investing.

7. You subscribed to 50 VC newsletters. You read maybe 3 of them.

8. You cannot tell if your sector is getting more investor attention or less. So you time your raise on a guess.

9. You spend two hours a week reading VC news. Next week you start over, and none of it gives enough insight.

Most founders we talk to hit five or six of these. None of it is a you problem.

What you actually want

• Walk into every meeting knowing what that investor wrote this month.

• A short list of the funds active in your sector, not a stale directory.

• One briefing that covers everything, instead of 50 unread newsletters.

• A read on whether investor attention on your sector is rising or falling, from data.

• Research you do not have to redo every week.

What our app does

Byblos Digital reads VC newsletters so you do not have to. It offers aggregated intel from 1,400+ investors and 60,000 newsletters in one place. Every piece is tagged by sector, theme, funding round, people move, and sentiment. Then it gives you what matters for your raise.

The features

One feed, followed VC. Every newsletter sits in one place. You read it here, not across 50 inboxes. You also do not have to hand over your email to each firm.

Intel. Any newsletter becomes tagged data. Funding rounds, people moves, thesis shifts. This is what powers everything above.

Filters. Sort the feed by fund, sector, theme, sentiment, content type, and other preferences. Find the investors active in your space in seconds.

VC Index cards. Open a card for any fund. See what they invest in, what they focus on, what they have been writing, and their links. Read this before you pitch.

Conviction tracking. See how much each sector is discussed over time, and whether the tone is positive or negative.

Watchlist. Build a list of your target funds. Track what they write during your raise. Fixes problems 1 and 4.

How to use it

1. Sign up. Pick your sectors and themes.

2. Follow the VCs you are targeting. Or let Byblos suggest them by sector.

3. Before any pitch, open that fund’s card. Read what they have written lately.

4. During your raise, keep your target funds on the watchlist and check what they write.

See it in action

Here is a short demo:

Want to make the batch early? How to get started:

🟣 Sign up over here and get a free 7-day trial of all features

Get free Byblos App

🟣 Get added to our Telegram group
🟣 Attend a 1:1 setup call with us
🟣 Receive free exclusive briefs from the 1,400+ VC database and use the app freely

Support us in launching on Product Hunt!

Follow our PH Launch

All of the Byblos Team: Lina, Ekin, Feri, Mert, Mike

Subscribe now

See what else came in today at byblos.digital →

Go premium. $10.99/month or $109/year, with a 7-day free trial. You get the full intelligence layer:

• Exclusive research from our database of 1,400 investors and 60,000 newsletters

• Weekly briefs on where capital is actually moving

• Direct 1:1 access to us

Get 7-day free trial

Nine VC firms wrote about the same narrow topic this week. How AI agents pay for things.

645 Ventures published two pieces on it. Activant Capital published the third part of a series. Animoca Brands launched a new research letter dedicated to it. The Algorand Foundation posted about it four times in six days. Archetype, Chainlink, Gradient Ventures, Village Global and Jump Capital all wrote about it too.

Our claim, before the evidence: the fight everyone is watching is which payment standard wins, and a startup cannot win that fight. The money is in two layers nobody is defending.

Quick thing before you dive in.

We launch our app on Product Hunt on the 30th of July.

It would mean a lot if you showed up and interacted with it on the day — a comment or an upvote goes a long way for a small team.

What you can do:

→ Follow every VC in your sector in one subscription
→ Filter your feed by fund, sector, theme, and sentiment
→ Open a VC card for any fund to see their investments, focus, and links
→ Build a watchlist around your raise and track what your target investors are writing

Want to make the batch early? How to get started:

🟣 Sign up over here and get a free 7-day trial of all features

Get free Byblos App

🟣 Get added to our Telegram group
🟣 Attend a 1:1 setup call with us
🟣 Receive free exclusive briefs from the 1,400+ VC database and use the app freely

There is no agreed way for an agent to pay for something.

An agent is software that completes a task without asking permission at each step. To finish most tasks it has to buy something. A dataset, an API call, a flight, a bag of coffee.

There is no standard way for it to do that. OpenAI, Google and Anthropic have each shipped their own payment protocol, and 645 Ventures reported this week that all three compete directly. A fourth standard called x402 is also live. It works by reusing HTTP 402, a status code meaning "Payment Required" that has existed since the 1990s and was never used for anything.

So a developer building an agent today picks one of four and hopes it wins.

Card networks do not fix this. Algorand made the point plainly. Cards were built for a person standing at a register. They were stretched to cover a browser tab, and that took decades of fraud tooling and extra checkout steps. An agent is not a person at a register. It can decide in milliseconds to call a hundred paid services on behalf of one user.

The evidence, part one.

The market is large and the best parts already have owners.

645 Ventures cites McKinsey projecting $3 to $5 trillion in agentic commerce volume by 2030. They also say the category moved from concept into early production over the past eighteen months. So there is real usage today, just small.

Then they say the part that matters for founders. The rails, the protocols and the identity layers belong to incumbents. Those are their words. Two layers are open to a startup: middleware and post-purchase.

Activant Capital listed the four problems that have to be solved before any of this works. Access, identity, settlement, governance.

In plain terms. Access means the agent is allowed to buy from this merchant. Identity means the merchant knows who is really behind the purchase. Settlement means the money actually moved and both sides agree it moved. Governance means somebody set rules for what the agent can spend, and can prove those rules were followed.

Activant says near-term capital is going into those four problems.

The evidence, part two.

Algorand is paying developers to use a standard.

A payment standard is worth nothing until people use it. Algorand is treating that as the real problem.

The Algorand Foundation put $100,000 and 500,000 ALGO into a public challenge for x402. It is running now. The task is specific: launch a paid API endpoint on Algorand mainnet, get real usage through the GoPlausible x402 facilitator, and climb a public leaderboard. Ten finalists present at Devcon 8 in India.

The entry rule is the part worth noting. A project does not count as entered until at least one real payment has settled on mainnet. Not a demo. A settled payment.

Animoca Brands and The Sandbox are running a smaller version of the same thing, a four-week agentic AI competition in Hong Kong with $10,000 in prizes, starting 28 July.

When the people building a standard start buying usage, usage is what the standard is missing.

Don't build a fifth standard.

OpenAI, Google and Anthropic own their protocols, and they own the models the agents run on. Algorand is spending cash to get x402 adopted. A startup does not outspend that.

The useful position is the opposite one. Merchants will end up accepting several standards at once, because their customers' agents will arrive using different ones. Build the layer that works regardless of which standard wins. That is the middleware layer 645 Ventures named.

The second open layer is everything after the purchase. Four firms described that gap this week without naming it as a company.

Archetype, announcing its investment in Nava, put it clearly. You would not hire a contractor for design work and then hand over your books, your passwords and a company credit card. That is what giving an agent payment access looks like right now.

Aaron Levie, speaking to Village Global, said he cannot hold an agent accountable to anything, and that turning an agent off means nothing to it. He also said agents will hit enterprise systems at 100 times the volume of human users.

Gradient Ventures asked how you back up a company running thousands of agents. They point out that enterprise backup and recovery is a market with $30 billion or more in annual revenue, built over 45 years, for a world where humans made the changes.

Animoca's new research letter says the gap between compute cost and agent revenue will decide which agent companies survive.

Put those together and the missing company is specific. A system that records what an agent bought, proves it was allowed to, and lets a business reverse it.

Three companies to build.

A merchant-side adapter. A merchant accepts an agent paying through any of the four standards without writing four integrations. This handles access and identity, two of Activant's four problems. Sell to merchants, not to agent developers.

Spend controls. A business sets rules for what each agent can buy, how much, and from whom. The system blocks everything else and keeps a record. This is Activant's governance problem, and it is what Levie means when he says he cannot hold an agent accountable.

Reconciliation and reversal. An agent makes 400 purchases in a month. Finance needs to know what each one was for and needs to reverse the wrong ones. This is the post-purchase layer 645 Ventures named and the problem Gradient described.

All three sell to the same customer.

The two objections

The first objection is timing. Agentic commerce revenue today is close to zero. The $3 to $5 trillion figure is a 2030 projection from McKinsey, cited by an investor with money in the category. Your first customers will be companies running agent pilots, not companies running agent operations. Plan for revenue in 2028.

The second objection is that incumbents extend into middleware themselves. OpenAI, Google or Anthropic could ship a merchant adapter. The reason none of them has is that each one wants its own standard to win, so none is motivated to make all four work equally well.

That gives you the thing to watch. If one standard takes clear majority usage, our conclusion stops holding, because a merchant only needs one integration and the adapter has no customer. Watch the Algorand leaderboard and watch whether OpenAI, Google and Anthropic start supporting each other's protocols. Either would be the signal.

Everything in today’s digest came from the Byblos live feed. The full newsletter archive, the VC directory, and tomorrow’s signal are all waiting for you at Byblos.digital - the investor interned, curated.

Get all the Byblos Links

That’s the data. Now go build something.

— Byblos.digital

Subscribe now

All of this was pulled from the Byblos live feed - 1,400+ VC firms, 60,000+ newsletters, one place.

See what else came in today at byblos.digital →

Go premium. $10.99/month or $109/year, with a 7-day free trial. You get the full intelligence layer:

• Exclusive research from our database of 1,400 investors and 60,000 newsletters

• Weekly briefs on where capital is actually moving

• Direct 1:1 access to us

Get 7-day free trial

Why Stablecoins, why now

That was not a coincidence. It is one of the loudest things investors are writing about right now.

Stocks, Treasuries, and bank deposits are being reissued as tokens. Stablecoins have become the layer that payment firms and banks want to own. The GENIUS Act gave US issuers a legal framework. All this influences the money flow.

So we pulled the list of every investor who actually writes about stablecoins, ranked by how much they cover it, from our database of 1,400 entries.

If you are building in stablecoins and raising, this is your target list. If you are tracking the space, this is who is paying attention.

Quick thing before you dive in.

We launch our app on Product Hunt in a week.

It would mean a lot if you showed up and interacted with it on the day — a comment or an upvote goes a long way for a small team.

What you can do:
→ Follow every VC in your sector in one subscription
→ Filter your feed by fund, sector, theme, and sentiment
→ Open a VC card for any fund to see their investments, focus, and links
→ Build a watchlist around your raise and track what your target investors are writing

Want to make the batch early? How to get started:

🟣 Sign up over here and get free access to all features until August 31

Get free Byblos App

🟣 Get added to our Telegram group
🟣 Attend a 1:1 setup call with us
🟣 Receive free exclusive briefs from the 1,400+ VC database and use the app freely

What’s in the list

Two sheets and a methodology tab.

Sheet 1 — Stablecoin Investors. 172 firms, ranked. For each firm you get:

• Interest tier. How much they cover stablecoins.

• Country, region, type, stage focus, AUM tier. The basics for filtering to firms that fit your round.

• Stablecoin pieces, last 90 days, weeks active, first and last covered. How much they write, and whether they are still active. A firm with a recent “last covered” date is warm.

• Top stablecoin topics. What they actually write about, with counts. This tells the angle.

• Top sectors. Their overall coverage profile, not just stablecoins. Shows you whether stablecoins are their core or a side interest.

• Notable deals reported. The top three stablecoin deals from their coverage. The full deal set is on sheet two.

• Fund vehicle news. New funds they raised or reported on. A firm that just closed a fund has fresh capital to deploy.

• Website and LinkedIn.

Sheet 2 — Stablecoin Deals. 271 deals, one per row. Every funding, M&A, and IPO event pulled from these firms’ stablecoin coverage, filtered to stablecoin-relevant deals only. Each deal shows how many firms reported it and who. The number of firms covering a deal is itself a signal — the more firms that wrote about it, the more the space cares.

Sheet 3 — Methodology. How we built it, what the tags mean, and the honest caveats.

How to use it

If you are raising. Filter sheet one to High tier, then to your region and stage. Check “last 90 days” so you only target firms still active in the space. Read their top topics to match your pitch to their angle. Open with what they actually wrote, using the notable deals column. Then go to the website and LinkedIn columns.

If you are tracking the market. Start with sheet two. Sort by “firms reporting” to see which deals the most investors cared about. That is your conviction map for where stablecoin attention is concentrating.

If you want the fast read. The top 20 rows of sheet one are the firms writing most about stablecoins, period.

Get the full list of 170+ Investors Interested in Stablecoins👇🏾👇🏾👇🏾

Read more

See what else came in today at byblos.digital →

Go premium. $10.99/month or $109/year, with a 7-day free trial. You get the full intelligence layer:

• Exclusive research from our database of 1,400 investors and 60,000 newsletters

• Weekly briefs on where capital is actually moving

• Direct 1:1 access to us

Get 7-day free trial

The outlook

This week we read 298 pieces from 159 firms. Here is what investors wrote about, the deals behind it, and who to talk to.

How to read the documents

Document 1 — The Weekly Brief

What it is. It covers the main narratives investors wrote about this week. Each story has a plain summary, the firms involved, and the deals that back it.

How we build it. We read every newsletter and blog from the funds we track for the week. We group what they wrote into narratives. We tag each narrative as NEW or CONTINUING.

What lifecycle means. A narrative starts when multiple firms write about it. As the number of publishers increases, it grows and becomes consensus, to then fade away.

How to use it. Read it to see what investors are paying attention to right now. Watch the NEW tags to catch themes before they are consensus. Watch the CONTINUING counts to see which themes last and which ones fade.

Document 2 — The Fund Activity Report

What it is. One card per fund. Each card shows what a fund wrote this week, what they wrote over the last three months, and how to talk to them. Look up the funds you care about.

How we build it. For each fund, we pull everything they published this week. Baseline is a fund's normal posting rate. Louder means they posted more than that this week. A fund getting louder on a theme is a fund paying more attention to it.

How to use it. Find the fund you want to pitch. Read their three-month focus and their talking points. Open your outreach with what they actually wrote. This is the difference between a cold pitch and a relevant one.

Read more

See what else came in today at byblos.digital →

Quick basics. A stablecoin is a crypto token built to stay worth about a dollar. The US now has a law for them, the GENIUS Act, one year old on July 18.

Total supply is just over $300 billion. USDT is about $184 billion. USDC is about $73 billion. Those two are most of the market.

So the coin layer is crowded. This week, almost nobody raised money to launch a new coin. They raised to build software that moves existing ones. Velocity raised $38 million for treasury tools. Cyclops raised $20 million selling payment software to payment companies.

For founders: the coin layer is crowded. The software layer is where the money is moving.

I’m bringing this up in the weekly article because our team is preparing a list of investors interested in stablecoin niche.

It will include all appropriate links, stage focus, fit notes, top topics covered by them in the last 6 months, rounds invested, and any recent important news.

Stay tuned for it, as well as for our incoming beta app launch!

We will be giving out free access to our app to users, setting them up, along with providing some tailored researches from the database with 800+ VCs.

What got funded

Three rounds closed this week. All three build tools on top of stablecoins that already exist.

Pact Labs puts USAT, Tether's US dollar stablecoin, into company payroll. That means paying staff in the stablecoin and letting them draw earned wages before payday. Pact says it has moved about $2 billion in loans and has 500,000 users. Velocity handles treasury and settlement for companies. Treasury means where a company keeps its cash. Settlement means moving money between two parties and confirming it is done. Velocity does this with stablecoins instead of banks, and it is run by Eric Queathem, who spent nine years at Worldpay. ADI Chain builds the rails for a dirham stablecoin backed by named UAE institutions.

The pattern: money went to payroll, treasury, settlement, and local-currency rails — the jobs stablecoins do — not to new stablecoins.

What to build

Payroll and early wage access. Workers get paid every two weeks, but bills come sooner. Pact Labs raised $7M, Tether-led, to run payroll on a stablecoin and let staff draw earned wages early. The startup: payroll on a regulated stablecoin, with early wage access built in.

Business treasury and settlement. Paying a supplier abroad through banks takes days and costs fees. With stablecoins it settles in minutes. Velocity raised $38M to do this for enterprises. The startup: a treasury tool to hold, move, and reconcile company cash in stablecoins.

Local-currency stablecoins with real backing. Most stablecoins are US dollars. ADI Chain raised $50M for a UAE dirham version backed by named local institutions. The startup: a stablecoin for one country's currency, with named regulated backers and one clear use.

These are directions the money is pointing, not guaranteed businesses. A funded competitor proves the problem is real. It also means you already have a competitor.

Which VCs to pitch, and how

Pitch each firm from the angle it just backed. Tether wants stablecoin distribution into real payments, so pitch payroll, wages, or Bitcoin rails. Dragonfly and FirstMark backed enterprise stablecoin infrastructure, so pitch treasury and settlement for companies. Coinbase Ventures backs anything that grows onchain payments. Ripple backs cross-border and B2B payments, and pushes its own RLUSD stablecoin, so pitch settlement and treasury.

One warning about the raw data. The feed lists firms like Lateral Capital, CL207, and White Star Capital next to these raises. Those are newsletters that reported the raise. Reporting a raise is not the same as leading it. Pitch the named lead investors above, not the newsletter that covered the news.

Timing

The GENIUS Act was signed July 18, 2025. It turned one year old on July 18, 2026. The law told US regulators to write the detailed rules by that date. They missed it. Only draft proposals exist. Comment periods run past July 18. There is a backstop date of January 18, 2027.

For a founder, the timing is useful. The law is passed, so the direction is set. The rules are not final, so the market is still forming. That window is a reason to raise and build now, before the rulebook locks in early 2027.

Discipline line: the law is real and passed. The detailed rulebook is not final yet. Do not tell an investor stablecoins are "fully regulated." They are legal. The rules are still being written. And big firms can lose interest in crypto fast — they have before. This is a pattern in the data, not a promise.

Everything in today’s digest came from the Byblos live feed. The full newsletter archive, the VC directory, and tomorrow’s signal are all waiting for you at Byblos.digital - the investor interned, curated.

Our X (for timely updates and founder/investor opps)

Our Linkedin (for connections and professional insights)

Our app (for VC index cards, the feed of newsletters, and upcoming AI summaries)

That’s the data. Now go build something.

— Byblos.digital

Subscribe now

The week's biggest shift is traditional finance moving onto crypto rails. The clearest signal was Open USD we mentioned before. It landed alongside Vanguard hiring its first head of digital assets, Circle winning federal banking approval, and tokenized equities reaching $3.86 billion in monthly volume, up 145% in a month. Funds keep making the same point: value is accruing to the settlement plumbing, not the tokens. The same week, Strategy sold roughly 3,588 Bitcoin for about $216 million to shore up dollar reserves, ending its "never sell" stance.

Read more

See what else came in today at byblos.digital →

This week’s headlines made tokens look strong. Robinhood launched a blockchain with tokenized stocks and perpetual futures. A perpetual future is a bet on a price with no end date. One trader turned $800 into $1 million on that chain in a few days. Memecoins traded heavily again. Bitcoin went up.

We looked at every partnership announced in the last ten days. 714 were between crypto companies only. 88 included a large mainstream company, like a bank, a card network, or a big consumer app. So about 9 in 10 crypto partnerships this week stayed inside crypto. The memecoin trading is part of that 9. It is crypto companies and crypto traders dealing with each other.

One caveat. The 714 number is inflated by older posts that got re-published this week. Treat it as rough. The gap between 714 and 88 is still real.

The deals that reached normal people came from the 88. Robinhood is a stock app with 27 million users, so its blockchain reached those users on day one. Coinbase’s x402 had its biggest month. x402 lets one piece of software pay another automatically using stablecoins. The companies using it this week were Amazon, Solana, and Exa. USDC reached 70% of stablecoin trading volume. That is people using it to move money.

What got funded this week

The biggest checks went to infrastructure, not to tokens. Infrastructure means the tools other products run on: compute, payments, model serving. Paradigm is a crypto-native fund. It raised $1.2 billion and said it will also invest in AI.

The same pattern holds over six months

Across the last six months, the crypto categories with the most funding rounds are the infrastructure ones. Real-world asset tokenization had 304 rounds. Stablecoin payments had 235. Stablecoins had 221. DeFi had 175. Bitcoin mining had 7. New token projects and Bitcoin mining are at the bottom.

Crypto is also not where most venture money goes right now. AI takes most of it. AI agents alone had 863 funding rounds in six months. That is more than any single crypto category.

What the data says

Crypto had an active week, but the activity and the growth are two different things. The activity was in tokens: memecoins, a new chain, a $1 million trade. The growth was in payment tools and in products that already have users. USDC moved money. Robinhood reached 27 million people. x402 got Amazon. The funding followed the same line, this week and across six months: the money is in payments, tokenization, and stablecoins, not in new tokens.

This is one week of headlines on top of six months of funding data. The six months is the solid part. Big companies can also stop caring about crypto quickly, and they have before. This is a pattern in the data, not a forecast.

The proof

• 714 crypto-only partnerships against 88 with a large mainstream company, last ten days. Rough count, inflated by re-published posts. The gap is clear.

• Biggest raises this week: Together AI $800M, Baseten $1.5B, Anysphere $900M, Fireworks $250M. All infrastructure. Paradigm closed a $1.2B fund and is expanding into AI.

• USDC reached 70% of stablecoin volume. Robinhood's chain reached its 27 million users.

• Six-month funding rounds: RWA tokenization 304, stablecoin payments 235, stablecoins 221, DeFi 175, Bitcoin mining 7.

• AI agents had 863 funding rounds in six months. That is more than any single crypto category.

Method: Byblos ingests the published output of 500+ Web3 VC firms and research desks into a tagged database. Deal and funding counts come from that database. Partnership counts cover the last ten days. Funding-by-category counts cover the last six months. Company figures were checked against public announcements where available. Two figures (USDC share, Robinhood user count) come from newsletter mentions and are worth a second check before quoting. Not investment advice.

Everything in today’s digest came from the Byblos live feed. The full newsletter archive, the VC directory, and tomorrow’s signal are all waiting for you at Byblos.digital - the investor interned, curated.

Our X (for timely updates and founder/investor opps)

Our Linkedin (for connections and professional insights)

Our app (for VC index cards, the feed of newsletters, and upcoming AI summaries)

That’s the data. Now go build something.

— Byblos.digital

Subscribe now

See what else came in today at byblos.digital →

Read more