Pillar
Releases, CVEs, events, partnerships
Cadence: real-time
Naive-N0.5-Flash, an open-weights coding model, claims 2,100 tokens per second, and an early hands-on review says its code holds up against DeepSeek. If that holds, agent builders can choose speed without giving up much quality.
Anthropic shipped Claude Fable 5.2 on IPO roadshow day with a 20% per-token price cut and a bigger appetite for tokens. For agent operators, cost per finished task now matters more than the rate card.
The most consequential line in Claude Code v2.1.318 is a log entry: a record of your answer when the agent asks to enable modifications it wrote itself. It shows where agent risk is moving.
A community demonstration of 'Weight Space Programming' shows models reading and editing neural network weights directly. That removes the training run from the self-improvement loop, and agent governance has nothing pointed at it.
Anthropic's S-1 reframes the AI race: 48% of revenue through cloud partners, an 85% margin, and a first adjusted-profitable quarter suggest agent labor has found a durable price. The open question is who owns the customer.
Ben Thompson's memory of a WordPress install that took far longer than five minutes is a map of where AI agents sit today. Setup friction filters users, the plumbing is turning boring, and the value is moving to whoever owns the relationship.
Moonshot's 2.8T Kimi K3.1 Pro beat Fable 5.1 ahead of a year-end prediction, priced at half of Claude Opus 5.5. For agent users, model choice is now a three-way trade between cost, speed, and capability.
OpenAI's Patch the Planet program co-discovered PHP flaws affecting 70% of the web. The real story is that two teams using the same model found the same bug, and the human disclosure pipeline was the part that leaked.
Modal hit $1B ARR and a $16B valuation, tripling both usage and valuation in five months. The growth came from sandboxes for AI-written code. Investors appear to be pricing execution infrastructure as a defensible layer of the agent stack, separate from intelligence.
Anthropic's new dynamic workflows, or "flows," treat multi-agent coordination as a first-class product inside Claude Managed Agents. Here is what that changes for teams running real workloads, and the failure points you'll want to plan for first.
Simon Willison asked ChatGPT to poll a GitHub repo hourly until Python 3.15 appeared. The small request shows where agents earn their keep: watching the lag between an upstream release and the moment your own tools can use it.
Anthropic priced Claude Haiku 5.5 to match OpenAI's GPT-6 Luna, and early coverage calls it better. That moves the competitive line in the tier where most agent labor actually runs.
Claude Code v2.1.294 fixed hooks that were allowing what they should block. The bug is patched. The deeper lesson for anyone relying on natural-language guardrails is not.
OpenAI's internal Navier-Stokes model reportedly produced 722 papers addressing 90 of math's top 500 open problems. The model was never the scarce part. Verification is, and so far it runs on one man's 24 years.
OpenAI priced GPT-6.1 Sol at $2/$10 per million tokens against Astra's $10/$50, while claiming wins on agentic benchmarks. The price is the story: a volume-first bet on agent workloads that should change how you budget reasoning-heavy tasks.
DevDay 2026 folded voice, documents, models, agents, and a marketplace into one ChatGPT account. That consolidation is convenient, and it makes every OpenAI login a far larger attack surface than any open harness users assemble themselves.
Anthropic's Sonnet 5.5 claims 30%+ faster runs and up to 30% lower costs at an unchanged price. Alongside the week's other launches, it signals that frontier models now compete on efficiency, and agent operators are the ones who benefit.
A closed launch drew 36 million views and six functional clones inside 48 hours. The architecture was never the moat, and the people who run agents should read the copy wave as a supply-chain warning, not a bargain bin.
Claude Cowork and chat are now one Claude, and the app keeps working after you close your laptop. The convenience story is real. The story nobody is telling is that the session boundary was doing security work, and it is gone.
OpenAI's biggest launch ever isn't a story about model specs. It's the moment the industry admitted the model layer no longer bottlenecks what your agent can do. The constraint moved to deployment, policy, and trust.
Greg Brockman is the face of OpenAI's Astra launch. The strategic tell isn't the model's raw capability. It's the growing case for alignment and controllability as the layer where value now accrues.
The new --restricted flag lets you strip an agent's ability to run commands or fetch the web before it starts. It's a small feature that flips the agent trust model from top-down to user-controlled.
OpenAI's Agents SDK now ships testing tools that let you validate agent workflows without a live model, sandbox, or network connection. That's not a convenience feature. It's an admission about who owns the layer that matters.
Google's Gemini 3.7 Flash reclaims ground it lost to Claude and GPT, reopening a three-way race for the model that powers the next generation of consumer agents.
