Sam Altman claims internal post-Astra model can solve problems the world's best mathematicians cannot
perksverse · x · 2026-09-17
Speaking with Salesforce CEO Marc Benioff, Sam Altman laid out a capability ladder for OpenAI's internal models:
- GPT-5.5: roughly as good as an average math professor
- GPT-5.6: top 1-2 percentile math professor
- Astra: a bit better than that
- An internal model past Astra: "can do things that the best mathematicians in the world cannot"
It's the most concrete framing yet of a post-Astra generation jump, implying research-grade mathematical capability in internal testing. The claim is an executive assertion without published evidence.
Related event: Altman says internal model beyond Astra surpasses top mathematicians(9 posts)→
More from Models
- Anthropic's server-side rule reignites the case for owning open models over renting AI — alex_verem · 2026-11-02
- Epoch AI: open-weight GLM 5.3 exploited the patch gap to hack 6 Firefox releases at ~$1,000 each — rohanpaul_ai · 2026-11-02
- SemiAnalysis: safety classifiers sandbag scores of Opus 5.5 and GPT-6 Astra — morqon · 2026-11-02
- Liquid AI's 1.7B LFM3 becomes first local model to nail live fixtures lookup on a MacBook — NielsRogge · 2026-11-02
- Claude upgrade-page traffic rebounded to 17.6M in September, near May's 18.6M peak — firstadopter · 2026-11-02
- Open-source GLiNER-PII scrubs 13 identifiers from a discharge note in 550ms on GPU — MaziyarPanahi · 2026-11-02