Posts

GPT-5.6 Sol Ultra is coming to OpenAI Codex. The confirmation came via a post on X from Tibo (@thsottiaux), and it’s triggering a wave of interest among developers who’ve been watching the GPT-5.6 family since its limited preview launched on June 26th. Here’s what you need to know about where things stand and why Sol Ultra in Codex matters for agentic coding workflows. The GPT-5.6 Family: A Quick Recap OpenAI previewed the GPT-5.6 series on June 26, 2026, restricted initially to trusted partners due to U.S. government coordination around frontier model rollouts. The family includes three tiers: ...

Enterprise AI governance just got a lot more concrete. Anthropic launched granular model access controls for Claude Enterprise on July 1, giving organization administrators the ability to restrict which Claude models employees can use — and cap how much reasoning effort they can apply per role. This is the first time Anthropic has exposed per-role, per-model RBAC including effort-level caps, making it a meaningful milestone for teams serious about governing agentic deployments. ...

If your agentic pipelines or LLM-powered CI/CD workflows rely on the GitHub Models API, you have 25 days to migrate. GitHub confirmed on July 1 that GitHub Models will be fully retired on July 30, 2026 — and that’s not a soft deprecation with a grace period. After July 30, the playground, model catalog, inference API, and bring-your-own-key (BYOK) endpoints go dark for everyone. Two planned brownouts will give you a preview of what “dark” looks like. Mark these dates: ...

Microsoft made its biggest structural bet on enterprise AI services yet this week with the launch of Microsoft Frontier Company on July 2. It’s a new operating unit backed by a $2.5 billion commitment and staffed with approximately 6,000 industry experts and forward-deployed AI engineers — embedded directly inside customer organizations to co-design, deploy, and govern AI systems at scale. The announcement signals something important: Microsoft doesn’t think enterprises can get from AI experimentation to production-scale deployment on their own. They’re building the scaffolding to do it for them. ...

If you’ve been watching OpenClaw’s release cadence, you know the team doesn’t slow down for summer. Today’s second July beta — 2026.7.1-beta.2 — lands with a wide-reaching feature set that touches model support, CLI tooling, messaging integrations, security scoping, and mobile apps all in one go. Let’s break down what’s new and why it matters for agentic AI practitioners. GPT-5.6 Family Now Supported The headliner for OpenAI users: GPT-5.6 is now a first-class model family in OpenClaw. Support covers three variants — gpt-5.6-sol, gpt-5.6-terra, and gpt-5.6-luna — across catalog, capability detection, and runtime selection paths. If you’ve been testing GPT-5.6 through other means, you can now route it natively through the Gateway without workaround configs. ...

Microsoft shipped dotnet-1.13.0 of the Agent Framework on July 3, 2026, bringing the Microsoft.Agents.AI NuGet package to version 1.13.0. This release focuses heavily on maturing the Skills API — graduating it from experimental to stable — along with meaningful Foundry Hosting improvements and new file editing capabilities. For .NET developers building production agents and multi-agent systems on the Microsoft stack, this is a noteworthy maintenance release with enough new capability to warrant an upgrade review, though the breaking changes require attention before you update. ...

Formal verification has long been the territory of academic mathematicians and safety-critical embedded systems engineers — deeply important work, but rarely the domain of general-purpose AI. Mistral is changing that narrative with Leanstral 1.5, a 119B Mixture-of-Experts model released under Apache-2.0 that doesn’t just write code in Lean 4 — it operates as a genuine agentic system, editing files and running bash commands to verify proofs. The benchmarks are remarkable. Leanstral 1.5 solves 587 out of 672 PutnamBench problems (a benchmark of undergraduate competition mathematics problems), achieves an 87% score on FATE-H, and saturates miniF2F — a standard Lean formalization benchmark. More practically, it was deployed against 57 real-world repositories and found 5 previously unknown bugs. ...

The AI scaling story just got more complicated — in the best possible way. Shanghai AI Lab (InternScience) has open-sourced Agents-A1, a 35-billion-parameter Mixture-of-Experts model that matches or beats trillion-parameter systems on agentic benchmarks. The kicker? It does this not by throwing more parameters at the problem, but by doing something far more interesting: training on dramatically longer agentic task trajectories. The paper is titled “Scaling the Horizon, Not the Parameters” (arXiv:2606.30616). That subtitle is the entire thesis. ...

Claude Fable 5 Is Leaving Subscriptions July 7 — What You Need to Know Claude Fable 5 just returned from a 19-day export control pause on July 1. Now, three days after its restoration, Anthropic has announced that it will leave Claude subscription plans on July 7 — less than a week after coming back. This is not the same situation as the export control removal. That was a regulatory hold. This is a capacity decision. And according to Anthropic, it’s temporary. ...

OpenAI’s Landmark Study: 99.8% of Employees Now Route Work Through Agentic AI We’ve all heard the “agentic AI is the future” prediction so many times it’s become background noise. OpenAI’s new research paper — “The Shift to Agentic AI: Evidence from Codex” — is different. It doesn’t make predictions. It presents data. Real usage numbers from real workers across three groups: individual users, organizational users, and OpenAI’s own 3,000+ employees. ...