Opus climbs from ~1500 to ~1750 Elo learning chess while Astra regresses
inductionheads · x · 2026-10-04
Developer petergostev ran a continuous-learning benchmark: models play 200 games against Stockfish, taking notes and choosing difficulty, with engine-based cheating banned. Opus impressively improved from 1500 to 1750 Elo, while GPT-6 Astra regressed — winning 0 of 68 games at full strength. A possible explanation: Astra's native 272k context window in Codex may crack under accumulated notes and files; the author considers rerunning with 1M context, likely only available on GPT-6.1-Sol. Live tracking available on his site.
More from Models
- Jeremy Howard: Google shipped the best model ever made; OpenAI its worst, same day — hugobowne · 2026-11-20
- Polymarket puts 83% odds on Google having the best AI model by year-end — Polymarket · 2026-11-20
- One year later, model naming is still a mess: the gpt-5.1-codex-max anniversary — heyneighbor · 2026-11-20
- Developer slams Grok models as 'garbage', shares coding agent's repetitive trial-and-error log — feregri_no · 2026-11-20
- Drex Router cuts small-game build cost to $1.26 vs $13.44 on Claude Opus alone — testingcatalog · 2026-11-20
- OpenAI slammed for AI slop: 'Just hire photographers to RLHF imagegen for a day' — flowersslop · 2026-11-20