GPT-6.1 Sol
OpenAI·
LLMsflagshipcloud
Overview
Update to GPT-6 Sol released one week after it, at OpenAI DevDay 2026, at the same price: $2 / $10 per million input/output tokens, cached input $0.10. Available in the API as gpt-6.1-sol, in ChatGPT Work and in Codex; not in regular ChatGPT conversations at launch. 1.05M-token context, 128K max output. No GPT-6.1 Astra shipped alongside it.
Capabilities and innovations
Agentic codingComputer useTool use / function callingReasoning effort control (none to max)1.05M token context windowAPI: gpt-6.1-solFirst point release of the GPT-6 Sol tier
Benchmarks
| Benchmark | Score | Source |
|---|---|---|
| Humanity’s Last Exam (no tools) | 52.9% | third party HLE 52.9% no-tools, measured by Artificial Analysis at max effort (Intelligence Index v4.3.2 run) (accessed 2026-10-24). |
| Terminal-Bench 4.0 | 55.05% | independent Terminal-Bench 4.0 55.05% (± 1.82), measured by Vals AI at max effort, rank 7 of 54 (accessed 2026-10-24). Artificial Analysis measures 56.1 in its own run. 4.0 is its own generation and not comparable with the 2.x or 3.0 fields. |
| MMMU-Pro | 86% | third party MMMU-Pro 86.0%, measured by Artificial Analysis at max effort (accessed 2026-10-24). |
| ARC-AGI-1 | 98.5% | independent ARC-AGI-1 98.5% Semi-Private, verified by ARC Prize (XHigh and High effort; Max 96.5, Medium 95.5), results dated 2026-09-29 (accessed 2026-10-24). |
| ARC-AGI-2 | 94.2% | independent ARC-AGI-2 94.2% Semi-Private, verified by ARC Prize (Max effort; XHigh and High 91.7, Medium 86.7), results dated 2026-09-29 (accessed 2026-10-24). |
- GPQA Diamond: independent evaluation pending — Not in the launch coverage of OpenAI's table. Artificial Analysis Intelligence Index v4.3 no longer runs GPQA Diamond and Vals AI has archived it, so no independent value exists (checked 2026-10-24).
- SWE-bench Pro: not reported by the vendor — OpenAI reports DeepSWE v1.1 instead, as for GPT-6 Sol.
- MMLU-Pro: not reported by the vendor
- Terminal-Bench 2.x: not reported by the vendor — No Terminal-Bench 2.x run reported. Vals AI and Artificial Analysis measure 4.0 only, recorded in terminal_bench_4.
- Terminal-Bench 3.0: not reported by the vendor
- DeepSWE v1.1: independent evaluation pending — Launch coverage quotes OpenAI at 75.2 (high effort), with GPT-6 Sol at 68.8, which matches what OpenAI published for Sol, and one tracker lists 71.9. openai.com and the coverage pages were not reachable from this environment, and Datacurve's public v1.1 board carries no GPT-6.1 Sol entry, so the figure is not entered until it is read at the source (checked 2026-10-24).
API pricing
$2 per million input tokens, $10 per million output tokens (USD, cheapest OpenRouter endpoint, checked 2026-10-24)
Reliability
- omniscience_index
- 41.5 ()independent
Links
More from OpenAI
Data curated by AI Model Timeline. See the methodology for admission criteria, benchmark eras and source priority.