Anthropic says Fable 5 reproduced 5 of OpenAI’s 10 Astra math advances in 24 hours
EricBuess · x · 2026-08-04
Anthropic’s Fable 5 reportedly reproduced half of OpenAI’s ten Astra math advances in 24 hours
Anthropic mathematician Levent Alpöge says Fable 5 autonomously solved five of the ten Astra math results within a day, using only a generic prompt and no internet access.
The post argues that these breakthroughs may not hinge on a single secret model or carefully tuned prompt. It also notes that the results still need to be published and independently verified, but suggests research-level math may be turning into a repeatable frontier capability rather than a one-off milestone.
More from Models
- Model outputs are no less buggy, but they now take hours to spot — jsuarez · 2026-08-04
- ASCII art may be a better taste benchmark for frontier models than you think — weswinder · 2026-08-04
- Athena 4B beats GPT-5.6 and Claude on shopper-action prediction benchmark — rohanpaul_ai · 2026-08-04
- MiniMax model reportedly cuts MacBook Pro runtime from 2h19m to 49m — bookwormengr · 2026-08-04
- DeepSeek V4 Pro still looks early, even as results are already being reported — teortaxesTex · 2026-08-04
- DeepSeek V4 Flash beat GLM 5.2 and Kimi K3 on a multi-app agent benchmark — LimpComedian1317 · 2026-08-04