Is Day 17 gone? Did Tibo fall asleep and forget?
Model intelligence feed
Model intelligence.
Follow the changes. Understand what comes next.
Editor’s pickOther providers
Valve's Answer to AI Slop and Clone Games on Steam: We Won't Police Developers' Tools
At Steam Indie Connect in Korea, Valve staff addressed AI slop and clone games head-on: no gatekeeping on AI tools, and the best defense against clones is community and steady updates.
Latest stories337
Newest published217218219220221222223224225226227228
Why the Claude API Now Rejects temperature, and How to Fix the 400 Error
On October 21, 2026, Anthropic released Claude API version 2026-10-21. Under this version, a request returns a 400 error if it sends temperature, top_p or top_k (even at the default value), places a document block directly in an assistant message, or sends a text block with a non-empty citations array in a user message.
Claude APIRead story
How to Direct Characters, Camera and Pacing in Kling 4.0
Kling 4.0 launched October 15, 2026. Kuaishou says it generates up to 30 seconds natively in one pass, combines up to 15 reference assets and accepts up to 10 keyframes. This guide maps those controls to characters, camera moves and pacing, and lists what Kuaishou has not announced yet, including prices, plans and regions.
Kling 4.0Read story
OpenAI Will Require Approval for Routers and AI Gateways Starting Nov 20
OpenAI's updated Usage Policies, effective November 20, 2026, require model-access intermediaries such as model routers and AI gateways to get OpenAI's review and approval first. The same update spells out a ban on using OpenAI models or outputs to train substitute models.
OpenAI Usage PoliciesRead story
OpenAI Split One Brain-Data Analysis Across Thousands of GPT-Rosalind Agents and Got Leads to Test
On October 21, 2026, OpenAI, PerturbAI and Scripps Research said thousands of GPT-Rosalind agents, each handling one gene disruption, analyzed existing data from millions of mouse brain cells and linked it to scientific literature in 32 minutes. In OpenAI's test, a single agent cited 11 references at the largest workload versus more than 700 for the distributed agents, and blinded experts preferred the distributed reports 75% of the time. The leads, including a possible DCTN1 and cholesterol link, are hypotheses that still need experimental validation.
GPT-RosalindRead story
Recursive Releases Flint, an Open-Weights Coding Model Built to Run on a Laptop
Recursive released Flint, its first model, on October 22, 2026 (Beijing time). It is an Apache 2.0 open-weights mixture-of-experts model with 39.8 billion total and about 2.5 billion active parameters, built for coding agents. Recursive reports 54.9% on Terminal-Bench 2.1, 89.7% on SWE-bench Pro v2 and 74.2% on SWE-bench Verified. It is free on OpenRouter at launch.
FlintRead story
What Is Claude Duet? Who Can Get It and How the Waitlist Works
Anthropic opened a research preview of Claude Duet on October 21, 2026. It is a standalone app where Claude acts as a proactive agent running on Claude Opus 5.5. It is open to Pro and Max subscribers in the US through a waitlist, at no extra cost, and it shares the plan's usage limits. Background checks count toward those limits. Team and Enterprise plans can't use it yet.
Claude DuetRead story
Claude Code and Cursor Both Run Claude, but Pay for It in Very Different Ways
Claude Code is Anthropic's own coding agent, bundled with Claude subscriptions. Cursor is an AI editor and agent now sold as part of The X Subscription, with its own models included and Claude billed separately at API list price. Here is how prices, limits, models and agent features compare, based on official pages, and which kinds of users each one suits.
Claude CodeRead story
Anthropic Uses Claude to Shrink Code's Attack Surface, Finding 493 Issues in 25 Projects
Anthropic published research on October 20, 2026 on using Claude to map which code is reachable and gate what a deployment does not need to expose. Across 25 open-source projects totalling over five million lines, Claude found 493 issues, including 162 undocumented security-relevant configuration combinations, 94 of which silently turned off a security control. In OpenSSH, fixes worked out with the maintainers removed roughly a fifth of the code an unauthenticated client could reach.
ClaudeRead story
Best Coding Agents Right Now, Ranked by Artificial Analysis (October 2026)
On the Artificial Analysis Coding Agent Index v1.5, three setups share the top spot: Claude Code with Claude Sonnet 5.5 (max), Devin Fusion CLI with Claude Opus 5.5 (max) plus GPT-6 Luna (high), and Muse Code with Muse Wonder Full (max), all between 68.0 and 68.4. This page groups the full leaderboard into tiers, lists each agent's model and setting, and gives token-based API cost per task.
Coding AgentsRead story
AI Helped Find Critical PHP Flaws. Here’s What Website Owners Should Check
If your website runs on WordPress or another PHP application, this article deserves a closer look. OpenAI has described a coordinated response to critical PHP vulnerabilities, and the PHP project has released fixes. The immediate question for a site owner is whether the PHP environment actually serving their website has been patched.
PHPRead story
OpenAI Is Coming to Google Cloud, Giving Businesses Another Way to Buy Its Models
OpenAI and Google are model rivals, but that has not stopped them from teaming up on cloud distribution. For businesses already running on Google Cloud, the partnership could change how they buy and…
OpenAIRead story
Kimi Code updated? Check the skill name before reinstalling the plugin
A plugin skill that stops responding after an update can look like a broken installation. With Kimi Code CLI 2.3.0, there is a simpler possibility to check first: the name used to invoke it has…
Kimi CodeRead story
From the Webmaster
Personal experiences, thoughts and notes.