GPT-6.2-Astra and 6.2-Sol land with sharper template-following, docs and spreadsheet upgrades
neelajj · x · 2026-11-18
OpenAI's michelelwang announces 6.2-Astra and 6.2-Sol. Compared to 6-Astra, 6.2-Astra follows slide templates far more closely (down to fonts and vector elements), writes clearer slides and docs, and produces more accurate, better-formatted spreadsheets.
Side-by-side evals in the screenshots:
- Docs: 6.2 answers directly whether the experiment worked, with clear concise prose; the old version buried facts in vague hedging
- Sheets: on a SOC 2 evidence tracker task, 6.2 produces clean, readable formatting; 6-Astra had awkward row heights and confusing distinctions
- Slides: 6.2 recreates a reference slide's exact fonts and vector elements and pulls relevant Gmail context into the takeaway title; 6-Astra ignored the template
Caveat: content accuracy is still a limitation — the model sometimes highlights the wrong info or rephrases. Tip: provide examples of your own past work when generating new slides/sheets/docs.
More from Models
- AI daily: GPT-6.2 launch, NVIDIA revenue doubles, AI singer tops Korean charts — xiaohu · 2026-11-18
- Early GPT-6.2 Astra & Sol hands-on: delegating whole workflows 'feels like a step change' — reach_vb · 2026-11-18
- After September's GPT quality slump, users doubt 6.2 restores reliability — Beautiful_Cheetah789 · 2026-11-18
- Indie dev: skip GPT-6.2 hype, what I want is a better Antigravity harness — yihui_indie · 2026-11-18
- New Gemini Astra shows cool behavior in novelty search harness — generativist · 2026-11-18
- Stanford finds all 8 LLM judges favor their own medical answers, skewing benchmarks — rohanpaul_ai · 2026-11-18