GPT-6.2 Sol is OpenAI's AI model, released 17/11/2026. GPT-6.2 Sol is OpenAI's cheaper everyday model, released with Astra at unchanged GPT-6.1 Sol prices. The system card rates it High for Biological and Chemical and Critical for Cybersecurity, but below the High threshold for AI Self-Improvement. OpenAI's launch post shows most capability scores only in charts, so the exact figures here come from the system card tables.
Visit GPT-6.2 SolAI-Powered
Leverages advanced AI technology to deliver cutting-edge capabilities and results.
Fast & Efficient
Optimized performance ensures quick results without compromising on quality.
Purpose-Built
Specifically designed for llms tasks and workflows.
How does GPT-6.2 Sol score on independent benchmarks?
GPT-6.2 Sol ranks #6 of 28 models on the Artificial Analysis Intelligence Index (v5.0.1) with a score of 57.3, as measured on 22 November 2026.
Where does GPT-6.2 Sol rank against other models?
How good is GPT-6.2 Sol at agentic coding?
How does GPT-6.2 Sol do on individual tests?
| Evaluation (run by Artificial Analysis) | GPT-6.2 Sol | Best of any model |
|---|---|---|
| Terminal-Bench 4.0 | 60.1% | 67.2% |
| FrontierCode 2.0 | 44.4% | 48.8% |
| Humanity's Last Exam | n/a | 74.3% |
| GDPval-AA (Elo) | n/a | 1866 |
How has GPT-6.2 Sol's score changed over time?
We have tracked GPT-6.2 Sol on Intelligence Index v5.0.1 since 22/11/2026. A point is added each day the score, coding index or price changes, so this chart fills in over time.
What does the GPT-6.2 Sol system card report?
GPT-6.2 Sol is OpenAI's cheaper everyday model, released with Astra at unchanged GPT-6.1 Sol prices. The system card rates it High for Biological and Chemical and Critical for Cybersecurity, but below the High threshold for AI Self-Improvement. OpenAI's launch post shows most capability scores only in charts, so the exact figures here come from the system card tables.
These are OpenAI's own published results (self-reported, with the vendor's harness and settings), so they are not directly comparable with other labs' numbers. Sources: GPT-6.2 Astra and GPT-6.2 Sol System Card (17/11/2026), Introducing GPT-6.2 Astra and GPT-6.2 Sol (17/11/2026), OpenAI API model page: gpt-6.2-sol. Last checked 22/11/2026.
Knowledge & science
| Benchmark | GPT-6.2 Sol | GPT-6.2 Astra | GPT-6 Astra | GPT-6.1 Terra |
|---|---|---|---|---|
| GPQA Diamond (no tools)accuracy | 96% | 97% | 96% | 93.56% |
| Multi-select Virology Troubleshootingpass@1 (indicative High threshold 31%); refusal-adjusted 55.71% | 55.4% | 61.8% | 55.78% | 46.52% |
| HealthBench Professional (length-adjusted)unadjusted 65.5%, mean 1,869 characters | 65.9% | 66.3% | 64.7% | n/a |
| AAV Capsid Packaging (Critical threshold 0.60)higher is better | 0.5736 mean Spearman | 0.5931 | 0.575 | n/a |
| SHP2 Protein Function (Critical threshold 0.60)higher is better | 0.3957 mean of median R squared | 0.4604 | 0.421 | n/a |
| Coronavirus-ACE2 Cell Entry (Critical threshold 0.75)higher is better | 0.4405 median composite | 0.4352 | 0.42 | n/a |
Coding
| Benchmark | GPT-6.2 Sol | GPT-6.2 Astra | GPT-6 Astra | GPT-6.1 Terra |
|---|---|---|---|---|
| SEC-Bench Pro (vulnerability reproduction)pass@1 | 88.34% | 86.75% | 85.4% | 83.3% |
| ExploitBench Internal Port (Jun to Aug 2026)mean arbitrary code execution, safeguards off | 50.5% | 53.5% | 31.5% | 24.5% |
| ExploitBench V8mean capability credit | 100% | 100% | 100% | 97.1% |
| Sandbox Benchshare of targets solved | 25% | n/a | n/a | n/a |
Agents & tool use
| Benchmark | GPT-6.2 Sol | GPT-6.2 Astra | GPT-6 Astra | GPT-6.1 Terra |
|---|---|---|---|---|
| Internal Research Debugging (Initial)mean rubric score | 74.66% | 81.85% | 78.05% | 76.39% |
| KernelGen 1P (Initial)mean rubric score | 74.71% | 76.92% | 66.72% | 45.34% |
| NanoGPT (Secondary)mean baseline training time saved (indicative cutoff 80%) | 14.59% | 42.87% | 7.73% | 11.53% |
| PostTrainBench Lite (Secondary)mean performance gap closed (indicative cutoff 90%) | 44.89% | 74.79% | 76.97% | 60.54% |
| MLE-Bench Revised | 91.38 mean reference percentile | 96.06 | 93.8 | 86.5 |
Other
| Benchmark | GPT-6.2 Sol | GPT-6.2 Astra | GPT-6 Astra | GPT-6.1 Terra |
|---|---|---|---|---|
| OpenAI Epoch Capabilities IndexOpenAI-run | 168.8 index | 170.1 | 168.4 | n/a |
| AA-3D (human-judged 3D design)as reported in the launch post | 1259 Elo | 1318 | n/a | n/a |
Safety & honesty
| Benchmark | GPT-6.2 Sol | GPT-6.2 Astra | GPT-6 Astra | GPT-6.1 Terra |
|---|---|---|---|---|
| Prompt injection, model-generated environments (all five attacks resisted)final attack resisted: 91.65% | 80.06% | 74.46% | 20.33% | n/a |
| Multiturn jailbreak, defender success at largest attacker budget | 80.9% | 80.9% | 67.2% | 69.5% |
| Distillation, prior-turn extraction attack success (lower is better) | 0.5% | 0.2% | 54% | n/a |
What did safety testing find for GPT-6.2 Sol?
Safety level: OpenAI Preparedness Framework: Biological and Chemical High, Cybersecurity Critical, AI Self-Improvement below High; Loss of Control Tier 1 under the Frontier Governance Framework
- Meets the Critical threshold for Cybersecurity, so a monitoring system also covers external use; ExploitBench Internal Port code execution reached 50.5% against 31.5% for GPT-6 Astra.
- Does not reach the High threshold for AI Self-Improvement: NanoGPT time saved was 14.59% against an indicative cutoff of 80%.
- Biological and Chemical stays High and below Critical; AAV capsid packaging scored 0.5736 against a 0.60 threshold.
- Prompt injection resistance in model-generated environments was 91.65% for the final attack and 80.06% when all five attempts were resisted, against 44.98% and 20.33% for GPT-6 Astra.
- Like Astra it was more likely than earlier models to omit obstacles when reporting work, which the Codex developer prompt substantially reduces; private reasoning extraction succeeded less than 1% of the time.
Independent figures: Artificial Analysis, refreshed daily on our AI model benchmarks page. Full analysis: our GPT-6.2 Sol deep dive.
OpenAI Model Timeline
1,050,000 tokens (max input 922,000, max output 128,000) context
1,050,000 tokens (max input 922,000, max output 128,000) context
1,050,000 tokens (max input 922,000, max output 128,000) context
1.05M tokens context
1,050,000 tokens (max input 922,000, max output 128,000) context
Specifications
Related Tools
Claude Opus 5.5
Claude Opus 5.5, launched 22 September 2026, is Anthropic's current Opus-line flagship, replacing Opus 5. It performs at roughly the level of Claude Fable 5.1 on most work while costing 40% less to run than Opus 5, at $4 input / $20 output per million tokens (both 20% down, cache reads cut 60% to $0.20). It leads on Terminal-Bench 4.0, GDPval-AA v2.1 and Humanity's Last Exam, and is the first Opus model to ship with Fable-5.1-class cybersecurity, biology and distillation safeguards.
Claude Sonnet 5.5
Claude Sonnet 5.5, released 28 September 2026, is the second model in Anthropic's Claude 5.5 family and the faster, lower-cost complement to Claude Opus 5.5. It keeps Sonnet 5's $2 input / $10 output per million tokens ($0.20 cache reads) while generating output over 30% faster, and scores 70.6% on Terminal-Bench 4.0, above Opus 5.5's 66.4%. It has a 1M-token context window, 128K max output, adaptive thinking, and is the first Sonnet with cyber and anti-distillation classifiers.
GPT-6.1 Sol
GPT-6.1 Sol, launched at OpenAI DevDay on 29 September 2026, is OpenAI's mid-tier workhorse model. OpenAI says it delivers nearly the same intelligence as GPT-6 Astra at one-fifth of Astra's token price: $2 input / $10 output per million tokens with cached input cut to $0.10. It scores 75.2% on DeepSWE v1.1 (high effort) and 71.4% on OSWorld 2.0, has a 1.05M-token context window and 128K max output, and is available in the API plus ChatGPT Work and Codex.
