AI News

Daily AI News for Builders

The stories worth reading, for builders and indie hackers. Updated all day.

Mistral in Talks to Raise €3B at €20B Valuation

Mistral is in talks to raise about €3 billion at a roughly €20 billion valuation, Bloomberg reported June 12. That would roughly double Europe's flagship AI lab from its last round and fund a compute war against US and Chinese rivals.

Jun 13, 2026 · 3 min read

technical

Visa Lets ChatGPT Agents Shop and Pay for You

Visa has embedded its payment network into ChatGPT, so AI agents can complete purchases at any Visa merchant, not just suggest them. Announced June 11, it pushes agentic commerce from demos into real checkout, with Mastercard, Google, and Stripe chasing the same rails.

Jun 13, 2026 · 3 min read

technical

DiffusionGemma: Google's 857 tok/s Diffusion LLM

Simon Willison dug into DiffusionGemma, an experimental diffusion-based Gemini model Google briefly released that generated 857 tokens per second. His June 10 writeup shows why diffusion text models could reset latency expectations for local inference.

Jun 13, 2026 · 3 min read

technical

TRACE Method Sharpens Long-Horizon Agent Reasoning

TRACE, a June 5 paper, improves long-horizon LLM-agent reasoning by aggregating evidence across steps instead of judging each step alone. It reports an aggregate F1 of 0.713 and recall of 0.844, with the largest gains on tasks that require linking far-apart clues.

Jun 13, 2026 · 3 min read

technical

Yang's Nobile Mobile Offers Cash Back on Wireless Bills

Andrew Yang launched Noble Mobile, a mobile virtual network operator that charges significantly less than major carriers and gives customers money back if they use less data. This innovative model aims to combat rising costs and the economic impact of AI on wages and jobs. With thousands of customers and millions in revenue, Noble Mobile is proving its business viability by sharing profits with subscribers to keep them loyal. The company's approach could be a blueprint for future consumer-facing businesses focused on reducing expenses.

Jun 13, 2026 · 3 min read

technical

Bezos's Prometheus Raises $12B to Automate Engineering

Jeff Bezos's Prometheus raised $12B at a $41B valuation to build an 'artificial general engineer' for physical systems. The round drew JPMorgan, Goldman Sachs, and BlackRock, and follows a $6.2B launch raise in late 2025.

Jun 13, 2026 · 3 min read

Like this? There's more.

Get it daily. 7am. Free. Reads in 5 minutes.

Join 2,024 builders reading daily.

technical

Simon Willison Tests Claude Fable 5 Across Effort Tiers

Simon Willison spent hours testing Anthropic's new Claude Fable 5, comparing its five thinking-effort levels on token cost and output. His read: the top tier is strong but pricey for routine work.

Jun 13, 2026 · 3 min read

technical

Mistral in Talks to Raise €3B at €20B Valuation

Mistral is raising about €3 billion ($3.5B) at a roughly €20 billion valuation, nearly double its September price. Europe's flagship lab needs the cash to keep pace with far richer US and Chinese rivals.

Jun 13, 2026 · 3 min read

technical

OpenAI Acquires Ona to Run Codex Agents for Days

OpenAI is buying German cloud startup Ona so its Codex coding agent can keep working for hours or days, even after you close your laptop. Codex now has 5M+ weekly active users, up from 3M in April.

Jun 13, 2026 · 3 min read

technical

Oracle Hits $638B AI Backlog, Plans $40B Raise

Oracle's remaining performance obligations jumped 363% to $638 billion on AI cloud contracts, yet shares fell about 10%. The drop came after Oracle said it will raise roughly $40 billion to fund data centers.

Jun 13, 2026 · 3 min read

technical

Anthropic, TCS Bring Claude to Regulated Industries

Anthropic partnered with Tata Consultancy Services to deploy Claude across regulated sectors like finance, healthcare, and government. TCS will give Claude to 50,000 employees in 56 countries and build Claude-powered products for clients.

Jun 13, 2026 · 3 min read

technical

Cohere's North Mini Code: 30B Coding Agent on One H100

Cohere open-sourced North Mini Code, a 30B-parameter MoE coding model that runs on a single H100 under Apache 2.0. It ships a 256K context window and hit 210 tokens per second in Artificial Analysis testing.

Jun 13, 2026 · 3 min read

Page 1 of 190