TUE, OCTOBER 06, 2026
Independent · In‑Depth · Practitioner‑Tested
✎ General

Twenty-Seven Companies Took a Third of Global Venture Capital

AI startups raised $102 billion in Q3 2026 - 64% of all global venture funding - across a record 27 billion-dollar rounds, while Nvidia's $12.9 billion acquisition of Hugging Face led the quarter's exits.

By AIToolsRecap October 6, 2026 5 min read 65 views
Home › Articles › General › AI News, Oct 6: AI Took 64% of All Venture Capital

AI Took Nearly Two Thirds of Global Venture Capital

Crunchbase's Q3 2026 figures, and the concentration is the story:

MetricQ3 2026
Global venture funding$159 billion
AI share$102 billion — 64%
Billion-dollar rounds27 — a record
Previous records16 in Q2, 14 in Q1
Share taken by those 27Roughly one third of all global VC

Twenty-seven companies took a third of everything raised worldwide, in every sector, in a quarter. That is the number to carry around.

The quarter was also down 25% from Q2's $212 billion and up 53% year on year from Q3 2025's $104 billion. It is simultaneously the weakest quarter of 2026 and the strongest since Q2 2022 - which tells you how far the baseline has moved rather than that anything is cooling.

Largest rounds: Databricks and Safe Superintelligence at $5 billion each, with Crusoe, Moonshot AI, Mistral AI, Nscale, The Boring Company and Kling AI all above $3 billion.

Largest exit: Nvidia's acquisition of Hugging Face at $12.9 billion, the quarter's biggest M&A deal. The largest IPO was ChangXin Memory Technologies at $8.6 billion raised.

Set beside last week's round of private raises - Instinct at $1 billion and ElevenLabs revalued at $22 billion - the pattern is a small number of companies absorbing most of the available capital while everything else competes for the remainder.

Meta Reportedly Halved Its Claude Code Seats

The Information reports Meta cut Claude Code access from around 60,000 users to roughly 30,000. Single-sourced, and neither company has commented, so treat it accordingly.

If accurate it is worth noting for what it is not: 30,000 internal seats of a competitor's coding tool is still an enormous deployment. A halving reads as a cost or licensing decision rather than a verdict on the product, and the remaining number is the more interesting figure.

A $3,499 PC for Local Agents

Ghost launched the Core PC, built for running AI agents locally, at $3,499.

The timing is not accidental. Between Apple tightening macOS full disk access on 2 October, the Muse system prompt disclosures, and Wikimedia's account of agents probing public services, the argument for an agent that runs on hardware you own and a network you control has had a good fortnight. Whether $3,499 is the right price for that argument is a separate question.

Four Labs Testified in New York

Executives from Anthropic, OpenAI, Google and Meta appeared before the New York City Council at a hearing on existential AI risk.

A municipal council has no jurisdiction over frontier model development. What it has is subpoena power over companies operating in the city and a public record. The venue matters less than the fact that all four sent someone.

Also

  • llama.cpp v0.6.0 adds MTP speculative decoding and GLM-5.3 support - the local-inference stack continuing to absorb techniques from the hosted side within weeks.

The Thread

Twenty-seven companies took a third of global venture capital. Nvidia bought Hugging Face for $12.9 billion. Meta halved its seats of a rival's coding tool. Ghost is selling a $3,499 box so you do not have to use anybody's cloud.

Concentration at the top, and at the same time a small steady push to run things locally. Both are responses to the same fact, which is that a handful of companies now own the infrastructure everything else runs on.

Tags
AI NewsGenerative AINvidia2026
⚑

Spot an inaccuracy?

We verify facts before publishing and correct errors promptly. If something in this article is wrong or outdated, let us know.

Report an error →
💡 AI Tools prompts
Prompt Guide
Best Claude AI Prompts for SEO (2026) — Content, Technical, and Comparison SEO
Claude Sonnet 5 and Opus 5 are strong for SEO work that requires writing quality, structured analysis, and long-form content generation. With 1M context, Claude can analyse an entire site's content structure, compare competing pages, and write complete article drafts in one session. These prompts cover the full SEO workflow: keyword research synthesis, content briefs, on-page optimisation, meta descriptions, technical audit interpretation, and comparison content that ranks above AI Overviews.
Get Prompts →
Prompt Guide
Best ChatGPT Prompts for SEO (2026) — GPT-5.6 and Browse
ChatGPT with GPT-5.6 Sol and Browse enabled is a capable SEO research tool — it can search the live web, analyse SERP results, and synthesise content briefs in a single session. GPT-5.6 Terra at $2.50/M offers a cost-efficient option for high-volume SEO content generation. These prompts are optimised for ChatGPT Plus with Browse, the ChatGPT Work product for larger projects, and the OpenAI API with web_search tool enabled.
Get Prompts →
Prompt Guide
Best Claude Opus 5 and Sonnet 5 Prompts for Writing (2026)
Claude Opus 5 and Sonnet 5 consistently produce the highest-quality long-form writing of any AI model in July 2026 — a lead documented across writing benchmarks and user testing since Claude 3 Opus. With 1M context and 128K output on Opus 5, Claude can write book chapters, complete reports, and long-form content without truncating. Sonnet 5 at $2/$10/M (intro through August 31) is the best value writing model available. These prompts are optimised for claude.ai Pro/Max, Claude Cowork, and the API.
Get Prompts →