OpenAI Testing 'Astra' Multi-Agent Model, May Face US Gov Pre-Release Safety Eval
APPSO · wechat · 2026-08-01
OpenAI is testing a new model family codenamed 'Astra', designed to deploy multiple AI agents that can collaborate on complex projects for extended periods. Sam Altman has reportedly demoed the model to US policymakers.
The final naming remains uncertain, potentially launching as GPT-6 or within the GPT-5 family. Notably, Astra might be among the first frontier models subject to the US government's pre-release safety evaluations, meaning its launch timeline partially depends on the finalization of new regulatory frameworks.
Related event: OpenAI Tests Multi-Agent Model Astra, Demoed to US Officials(17 posts)→
More from Models
- ChatGPT co-inventor launches Jev: claims 20-200x faster, 40-400x cheaper than LLMs — GabGarrett · 2026-09-18
- Ran Jev across 10 services for hours, still couldn't spend $1 — multiply_matrix · 2026-09-18
- Tencent's Hy4 Preview ranks #4 among open-weight models, cheapest in top ten — mariofilhoml · 2026-09-18
- Simple letter-counting test exposes huge gap: GPT-6-Astra hits 93%, Fable 5.1 flounders — scaling01 · 2026-09-18
- Jev, a 'System One' model by Typesafe, launches on OpenRouter with typed decisions instead of text — majidmanzarpour · 2026-09-18
- Zhipu claims AI autonomously discovered a WeWorm-exploitable vulnerability — teortaxesTex · 2026-09-18