AI TrackerGet the app

AI news,
sorted for builders.

Launches, price changes and retirements from every AI provider, summarized. The same feed as the app's News tab.

xAI· Products

Grok Bot can now make phone calls

The Grok Bot account on X announced that Grok Bot can now make phone calls. The post drew about 1,300 likes and 430 replies within the first hour.Models

Artificial Analysis makes Time per Task the headline metric of its API performance benchmarks

Artificial Analysis announced a new version of its API performance benchmarking, which covers about 1,000 serverless endpoints it tests several times a day for speed, latency and caching. Time per Task, how long a model takes to finish a task, replaces streaming speed as the headline metric. Artificial Analysis divides a model's output tokens per Intelligence Index task by its End-to-End Output Speed.Models

Arena: on long-horizon Agent Arena tasks, Claude Opus 5.5 works longest and Kimi K3.1 Pro shortest

Arena, the leaderboard publisher, says long-horizon tasks in its Agent Arena, which run hundreds of turns (porting a codebase to Rust, auditing an investment strategy document, writing a math textbook), widen the differences between frontier models.Coding

Cloudflare open-sources a security-audit skill for coding agents

Cloudflare's developer account released cloudflare/security-audit-skill, an open-source skill that turns a coding agent into a team of security auditors. Cloudflare says the skill seeded its own vulnerability-discovery harness.OpenAI· Models

OpenAI brings GPT-5.6 Cyber to AWS GovCloud (US-East) through Daybreak Red

OpenAI says Daybreak Red, its access program for purpose-trained cybersecurity models, now offers GPT-5.6 Cyber to approved customers in AWS GovCloud (US-East) through Amazon Bedrock. It builds on the earlier availability of Daybreak models on Amazon Bedrock.Infrastructure

Cloudflare adds analytics for personal agents visiting your site, with finer access policies

Cloudflare says it is launching new analytics that show how personal agents access a site on behalf of humans, along with more granular access policies for acting on what the analytics show. The details are in a Cloudflare blog post titled "personal agents visibility and control".OpenAI· Coding· 3 discussions

OpenAI will acquire Pierre Computer, the code-storage startup behind Codex's file and diff views

Axios reports, citing OpenAI, that OpenAI will acquire Pierre Computer, a developer tools startup, as it moves into territory long dominated by Microsoft's GitHub. According to the report, OpenAI plans to fold Pierre's code storage technology into Codex, extending the coding tool into the infrastructure where developers store and manage code.Mistral AI· Models

Mistral shows Mistral Large 4 finding, reproducing and patching a Django CVE

Mistral posted a video on X of Mistral Large 4 ("Le Chonk") working on a recent Django CVE. In its description, 133 agents map the code, hunt for flaws, test attack scenarios and confirm the bug, and the model then writes a patch: triage, reproduction, fix. The video is sped up 90x.Anthropic· Models

Anthropic says its Claude Code game rule was also catching games it should have built

Addy Osmani, an Anthropic member of technical staff, posted an update on the rule added to the Opus 5.5 prompt in Claude Code to steer the model away from recreating specific existing games. He writes that last week some users saw Claude Code push back on game projects it should have simply built, and that the rule was also affecting such projects.xAI· Models

xAI retires grok-imagine-image-quality and grok-imagine-image-pro, serving them with a cheaper model

According to xAI's API release notes, requests to grok-imagine-image-quality and grok-imagine-image-pro are now served by grok-imagine-image-2.0-low, which is grok-imagine-image-2.0 with quality set to low, at a lower per-image price. xAI says no request changes are needed.Models

Agent Arena ranks Tencent's Hy4 #14 overall, at about 62% less cost per task than Kimi K3.1 Pro

Arena says Tencent's Hy4 landed at #14 overall on its Agent Arena, which scores models on real agentic sessions from its user community. Arena says the result draws on more than 8,500 long-horizon sessions and a net improvement of 4.7%, and that Hy4 joins the Agent Arena Pareto frontier of score against cost.Models· 5 discussions

Tencent releases Hy4, now on OpenRouter and Tencent Cloud

Tencent's Hy4 is live on OpenRouter, where it is listed as tencent/hy4 with a 1,048,576-token context window. The AI-model news account AiBattle reports that it is also on Tencent Cloud, and that Tencent plans to release the model's weights and is working on a smaller Hy4 Flash. Tencent Hunyuan's own announcement says Hy4 is new with vision, calls it "firmly among the leading open models" and says it was built with cyber and biosecurity in mind. Tencent's research page for the model is at hy.tencent.ai/research/hy4.OpenRouter· Models

OpenRouter's catalog gives three Inception Mercury models an expiration date of November 8

OpenRouter's model catalog now shows an expiration date of November 8, 2026 for three models from Inception: Mercury 3 (inception/mercury-3, 1,000,000-token context), Mercury 2.5 (inception/mercury-2.5, 260,000 tokens) and Mercury 2 (inception/mercury-2, 128,000 tokens). None of the entries had an expiration date before. This is a catalog field only; the entries do not say why or what happens to the models after that date.Anthropic· Coding

Claude Code 2.1.318 stops installing the IDE extension from IDE terminals and warns before a claude.ai sign-in expires

Claude Code v2.1.318, released November 3, 2026, per its release notes:Coding

Vercel CI now runs on GitHub repos that have no Vercel project

Vercel said its CI product, launched in beta on October 28, now works on any GitHub repository without a Vercel project. To use it, a user installs the Vercel GitHub App, links their GitHub account to Vercel, and runs vercel ci enable in the repository root.Infrastructure

Cloudflare D1 starts capping query results at 50 MB on November 4

Cloudflare's developer account says D1 begins enforcing a 50 MB limit on the size of a query's result on November 4. Cloudflare announced the change in its changelog on October 30.Coding

Better Auth ships a security update, v1.7.16 and v1.8.0-beta.5, and asks users to upgrade

Better Auth, the TypeScript authentication library, says on X that v1.7.16 and v1.8.0-beta.5 are on npm as its November 2026 security update, and tells users to upgrade to the fixed version of their release line. Its advisory is at better-auth.com/blog/security-release-2026-11-03.Anthropic· Coding

Cat Wu, a product lead on Claude Code, has left Anthropic

Cat Wu announced on X that her last day at Anthropic was Friday, October 30. She joined the Claude Code team about two years ago, when it was a small internal tool, and thanked the teams behind Claude Code, Cowork and Tag.Coding

Vercel launches Security Review in public beta

Vercel launched Security Review in public beta. It runs vulnerability reviews on every push, and Vercel Agent can ship fixes in one click.Coding

Vercel CLI adds `vercel doctor`, a setup check built for coding agents

Vercel's CLI now has a vercel doctor command that diagnoses setup problems so an agent can find and fix them. Vercel says it shows the fix command for each issue, runs without a login or a linked project, never prompts, logs in or links, and outputs structured results with --json.
Older news