Heavy user: GPT-6 best ever at math but tiny context window hobbles it
pvncher · x · 2026-09-06
A heavy user comparing extensive hands-on time with both models: GPT-6 Max is "the best model ever" for math/ML tasks, beating 5.2 Pro (a GPT-6 Pro tier has just appeared in the web interface). GPT-6's biggest weakness is its tiny context window — "like talking to a super-smart Dory". Fable 5.1 remains better at long agentic tasks, instruction following, and finding elegant coding solutions. The replier adds a tip: you can raise the context window in the Codex config — it consumes usage linearly, so mind the cache.
Related event: GPT-6 Tops Math but Small Context Hurts Long Tasks(4 posts)→
More from coding & agent
- Indie dev's agent tracks 519 directories, 311 submitted for SEO backlinks — fekdaoui · 2026-10-23
- ScienceClaw Hackathon at MIT Media Lab to build decentralized AI agent collectives for science — ProfBuehlerMIT · 2026-10-23
- What should a native agent harness handle beyond tool calling? — Glittering812 · 2026-10-23
- Running ~100 DeepSeek v4.1 flash agents in parallel for under $20/month — zack_overflow · 2026-10-23
- MCP's stateless 2026-07-28 spec drops session-id — how are stateful tools adapting? — Stock-Pumpkin-8859 · 2026-10-23
- Dev adds local MCP server to Mac screenshot app so Claude designs from your references — Embarrassed-Ebb-740 · 2026-10-23