GPT-6.1 Sol
GPT-6.1 Sol is a closed AI model from OpenAI with an index score of 191 (GPT-4 = 100), rank 9 of 217.
191175–207Index score#9 of 217 models
$2.0/ $10Price per million tokensinput / output · via OpenRouter
1.05MContext windowtokens it reads at once
29 Sep 2026Released2 months ago
Where it sitson the frontier
GPT-6 Astra reached this score 1 month earlier.
Test resultsby area · tick = best by any model
Reasoning & knowledgeevidence
Mathsevidence
Codingevidence
Agentsevidence
Professional workevidence
Novel problemsevidence
Long tasksno evidence yet
Third-party results only, at the model’s best standard setting. Hover a bar for who ran it.
What the lab says1 of 5 also run by third parties
| Test | OpenAI says | Third parties |
|---|---|---|
| Terminal-Bench Science 0.1max effort | 57.0% | 52.9%Vals AI |
| DeepSWE v1.1best of five effort levels | 75.2% | not run yet |
| GDP.pdfbest of five effort levels | 32.0% | not run yet |
| AutomationBench 1.0.6best of five effort levels | 36.1% | not run yet |
| OSWorld 2.0offline set, partial credit, best of five effort levels | 71.4% | not run yet |
From OpenAI’s launch post, Introducing GPT-6.1 Sol, taken from the data behind the post's charts. Shown, never scored; settings can differ from the third-party run.
What others sayrank on 12 leaderboards
The Model Indexthird-party tests, our method#9of 217 · 191
Epoch Capabilities Indexoverall capability from many benchmarks#6of 231 · 168.6LMArena · Codingcoding questions in chat#10of 357 · 1540LMArena · Mathsmaths questions in chat#17of 346 · 1499LMArena · Instruction followingfollowing detailed instructions#20of 362 · 1482LMArena · WebDevbuilding web apps and front ends#8of 133 · 1742LMArena · Visionunderstanding images#10of 145 · 1293LMArena · Textoverall chat quality, judged by people#27of 362 · 1481LMArena · Long promptslong, complex requests#28of 340 · 1488LMArena · Creative writingcreative writing#36of 360 · 1449LiveBench · Overalloverall, on regularly refreshed questions#7of 38 · 81.4LMArena · Searchanswering with web search#8of 41 · 1230Best attop three among current models
Where to use itprices checked 18 Nov 2026
API providers 3 · per 1M tokens, in / out
Host prices from OpenRouter; going direct can differ.
Facts and sources
Released29 Sep 2026
Announcementopenai.com ↗
Release entryGPT-6.1 Sol
Scored asGPT-6.1 Sol
Output speed–
Hugging Faceclosed weights