News archive
All posts and signals, newest first.
SearchGLM-5.2 spotted on OpenRouter
DeepSeek V4 Flash spotted on OpenRouter
AI may have just cracked one of the hardest open problems in mathematics: Navier–Stokes
And the story behind it is getting more interesting. Tristan Buckmaster says OpenAI told him its internal research model had generated a roughly…
DeepSeek V4 Flash 0731 spotted on OpenRouter
Qwen3 14B: Catalogue listing
DeepSeek-V4-Pro-0813 spotted on OpenRouter
CODEX RESET INCOMING
LA 6 PM • NYC 9 PM 🇧🇷 São Paulo 10 PM 🇬🇧 London 2 AM 🇩🇪 Berlin 3 AM 🇮🇳 India 6:30 AM 🇨🇳 Shanghai 9 AM 🇯🇵 Tokyo 10 AM
GPT-6 Astra is officially #1 on the updated Artificial Analysis Intelligence Index, sharing the top spot with Claude Fable 5.1
OpenClaw can use Claude subscriptions again
It works through Claude Code CLI, while most OpenClaw features remain available through MCP: browser, memory, sessions, subagents, cron jobs, etc.…
GPT 6 Astra set a new record in MazeBench, a 3D puzzle benchmark
It explored 40% of all rooms and collected 14% of gems Fable 5.1 reached just 4.3% of rooms and collected 2% of gems
Guys, I got accepted for monetization on X
This gives me even more motivation to keep pushing and make even better content for you. Thanks to everyone who sticks around and reads my little…
OpenAI agents went from behind humans to 3.1x in just a few months
Before June, agents were still doing less work than humans If this trend continues, that could be around 7x by the end of the year
GPT-6 Astra Fast mode quietly went from 1.5x to 2x speed, with the same 2.5x usage multiplier
It’s still pretty expensive though. A task that would count as 1,000 usage units on Standard would count as 2,500 on Fast, so that’s still 150%…
Correction: I got the calculation in this post wrong
A 2.5× usage multiplier means 150% more usage, not 25%. Corrected calculation: https://alexgetman.com/403/gpt-6-astra-fast-mode-quietly-went-from-1…
GPT-6 Sol is already being tested internally
There is a good chance it drops on DevDay, Sep 29. Tibo said Astra pulled some OpenAI releases forward by six months. Sol feels like the obvious…
GPT 6 Astra and Fable 5.1 got 20 attempts to pick up a block and place it into a bowl
Astra: 95% success, 2.5 min, $0.94 per run Fable 5.1: 40% success, 6.8 min, $2.12 per run Astra also used 6× fewer output tokens, just 2.1k per run.
Claude is no longer the best model for frontend
GPT-6 Astra just took the #1 spot, knocking Claude Fable 5.1 off the top The jump is massive: the previous GPT generation, GPT-5.6 Sol, was #13 A…
GPT 6 Astra and Fable 5.1 got 20 attempts to pick up a block and place it into a bowl
Astra: 95% success, 2.5 min, $0.94 per run Fable 5.1: 40% success, 6.8 min, $2.12 per run Astra also used 6× fewer output tokens, just 2.1k per run.
OpenAI quietly bumped Astra’s limits by 50% across all plans
On Plus 5-hour limit went from 3–30 messages to 5–45.
Plus users got access to GPT-6 Astra
GPT-6 Astra is out for Pro users
GPT-6 Astra looked like a generational leap on OpenAI’s benchmarks
Artificial Analysis puts it closer to old Fable 5 territory.
OpenAI will compensate paid ChatGPT users still waiting for Astra
You’ll get one banked reset for every day without access, starting today. The first one lands in about 3 hours.
GPT-6 Astra is absolutely crushing everyone on the benchmarks
Claude Code is having issues again
Last month: Claude Code: 3 major outages + 6 degraded incidents Codex: 1 degraded
GPT 6 appears in an official OpenAI repo
OpenAI opened a new branch with GPT 6 right in the name. This is the clearest confirmation yet that the next generation will be called GPT 6, and it…























