Reasoning
198-question Diamond subset; independent AA evaluation
- 01
- 02
- 03
- 03
- 03
GPT-5.6 Sol
Across the task benchmarks we track, GPT-5.6 Sol’s strongest relative results are in reasoning, coding, and knowledge work. On GPQA Diamond, it scores 94.1%, ranking joint #3 among 30 models with reported results in our dataset.
Updated: Data as of:
Strongest results
| Benchmark | Measures | Score |
|---|---|---|
| GPQA ↗Diamond / Artificial Analysis | Reasoning | 94.14%Joint #3 of 30 |
| Humanity's Last Exam ↗Stable HLE — text-only subset | Expert reasoning | 49.5%#8 of 68 |
| GDP.pdf ↗100 tasks / All-pass / five attempts | Professional document reasoning | 27.8%#7 of 55 |
| AutomationBench-AA ↗1.0.6 / AA private 657 | Work automations | 60.1%#17 of 55 |
Model details
| Model | GPT-5.6 Sol |
|---|---|
| Agent Index v1.2 | Insufficient coverage (3/7) · Preview |
| Creator | OpenAI (opens in a new tab) |
| Availability | Proprietary |
| Context size | 1,000,000 tokens (opens in a new tab) |
| Max output | 128,000 tokens (opens in a new tab) |
| Knowledge cutoff | Not reported |
| License | Not reported |
Pricing and performance
Per provider and reported configuration. Prices are USD per 1M tokens.
| Provider / configuration | Contexttokens | InputUSD / 1M tokens | OutputUSD / 1M tokens | LatencyTTFT · seconds | Speedtokens / sec |
|---|---|---|---|---|---|
| GPT-5.6 Sol (max) | 1,000,000 | $4.00 (opens in a new tab) | $20.00 (opens in a new tab) | 125.777 (opens in a new tab) | 120.006 (opens in a new tab) |
| GPT-5.6 Sol (max) | 524,288 | $4.00 (opens in a new tab) | $20.00 (opens in a new tab) | 89.145 (opens in a new tab) | 125.307 (opens in a new tab) |
| GPT-5.6 Sol (max) | 1,000,000 | $4.00 (opens in a new tab) | $20.00 (opens in a new tab) | 115.676 (opens in a new tab) | 126.629 (opens in a new tab) |
Benchmark performance
GPT-5.6 Sol in color, alongside leading models. Latest stored reports, best effort per report date.
GPQA Diamond ↗
Coding
113 tasks; mini-swe-agent
- 01
- 01
- 01
- 04
- 05
Terminal Use
Official 4.0 board; native agent harnesses
- 01
- 02
- 03
- 03
- 05
- 10
Web Research
Professional benchmark table; OpenAI Astra release comparison
- 01
- 02
- 03
- 04
Work Automations
Private held-out set; API mode; 50 steps; strict pass/fail
- 01
- 02
- 03
- 04
- 05
- 07
Knowledge Work
220 tasks; Stirrup agent; pairwise Elo
- 01
- 02
- 03
- 04
- 05
- 14
Reported benchmarks
All published results in our dataset, with versions, evaluation settings, and sources.
| Benchmark | Score | Evaluation setup | Source |
|---|---|---|---|
| AA-Briefcase1.1 / combined EloCombined Elo · Higher is better | 1,479 Elo | selection: best reported configuration per release · GPT-5.6 Sol (max) · tools: benchmark-defined | Artificial Analysis (opens in a new tab)independent |
| AutomationBench1.0.6 / Zapier held-out private setStrict task completion · Higher is better | 3.81% | GPT-5.6 Sol; private held-out 657 tasks; API mode; 50-step cap; strict completion | Zapier ↗ (opens in a new tab)benchmark publisher |
| AutomationBench1.0.6 / Zapier held-out private setStrict task completion · Higher is better | 11.72% | GPT-5.6 Sol (Low); private held-out 657 tasks; API mode; 50-step cap; strict completion | Zapier ↗ (opens in a new tab)benchmark publisher |
| AutomationBench1.0.6 / Zapier held-out private setStrict task completion · Higher is better | 19.63% | GPT-5.6 Sol (Medium); private held-out 657 tasks; API mode; 50-step cap; strict completion | Zapier ↗ (opens in a new tab)benchmark publisher |
| AutomationBench1.0.6 / Zapier held-out private setStrict task completion · Higher is better | 24.81% | GPT-5.6 Sol (High); private held-out 657 tasks; API mode; 50-step cap; strict completion | Zapier ↗ (opens in a new tab)benchmark publisher |
| AutomationBench1.0.6 / Zapier held-out private setStrict task completion · Higher is better | 26.33% | GPT-5.6 Sol (XHigh); private held-out 657 tasks; API mode; 50-step cap; strict completion | Zapier ↗ (opens in a new tab)benchmark publisher |
| AutomationBench1.0.6 / Zapier held-out private setStrict task completion · Higher is better | 28.77% | GPT-5.6 Sol (Max); private held-out 657 tasks; API mode; 50-step cap; strict completion | Zapier ↗ (opens in a new tab)benchmark publisher |
| AutomationBench-AA1.0.6 / AA private 657Objectives completed without guardrail violations · Higher is better | 60.1% | reported precision decimals: 1 · result quality: reported · GPT-5.6 Sol (max); private held-out 657 tasks; 50 turns; objective completion with zero credit for any guardrail violation. · source model href: /models/gpt-5-6-sol · task count: 657 · Tools enabled · turn cap: 50 | Artificial Analysis (opens in a new tab)independent |
| BrowseCompOpenAI Astra launch comparison / September 2026Accuracy · Higher is better | 90.4% | OpenAI GPT-6 Astra launch comparison; best reported reasoning setting; see source evaluation footnotes | OpenAI ↗ (opens in a new tab)provider reported |
| DeepSWEv1.1Pass@1 · Higher is better | 73% | mini-swe-agent; effort=max; ±3 percentage points | Datacurve ↗ (opens in a new tab)benchmark publisher |
| DeepSWEv1.1Pass@1 · Higher is better | 73% | max; ±3 percentage points | Datacurve ↗ (opens in a new tab)benchmark publisher |
| DeepSWEv1.1 / OpenAI Astra launch comparisonPass@1 · Higher is better | 72.7% | OpenAI GPT-6 Astra launch comparison; best reported reasoning setting; see source evaluation footnotes | OpenAI ↗ (opens in a new tab)provider reported |
| GDP.pdf100 tasks / All-pass / five attemptsAccuracy · Higher is better | 27.8% | selection: best reported configuration per release · GPT-5.6 Sol (high) · tools: benchmark-defined | Artificial Analysis (opens in a new tab)independent |
| GDPval-AA2.1Elo · Higher is better | 1,577.77 Elo | GPT-5.6 Sol (Xhigh); 220 tasks; Stirrup agent; pairwise Elo | Artificial Analysis (opens in a new tab)independent |
| GDPval-AA2.1Elo · Higher is better | 1,617.83 Elo | GPT-5.6 Sol (Max); 220 tasks; Stirrup agent; pairwise Elo | Artificial Analysis (opens in a new tab)independent |
| GDPval-AA2.1Elo · Higher is better | 1,511.8 Elo | GPT-5.6 Sol (High); 220 tasks; Stirrup agent; pairwise Elo | Artificial Analysis (opens in a new tab)independent |
| GDPval-AA2.1Elo · Higher is better | 1,249.91 Elo | GPT-5.6 Sol (Non-reasoning); 220 tasks; Stirrup agent; pairwise Elo | Artificial Analysis (opens in a new tab)independent |
| GDPval-AA2.1Elo · Higher is better | 1,431.69 Elo | GPT-5.6 Sol (Medium); 220 tasks; Stirrup agent; pairwise Elo | Artificial Analysis (opens in a new tab)independent |
| GDPval-AA2.1Elo · Higher is better | 1,312.98 Elo | GPT-5.6 Sol (Low); 220 tasks; Stirrup agent; pairwise Elo | Artificial Analysis (opens in a new tab)independent |
| GPQADiamond / Artificial AnalysisAccuracy · Higher is better | 93.131% | GPT-5.6 Sol (Xhigh); 198-question Diamond subset; independent AA evaluation | Artificial Analysis (opens in a new tab)independent |
| GPQADiamond / Artificial AnalysisAccuracy · Higher is better | 94.141% | GPT-5.6 Sol (Max); 198-question Diamond subset; independent AA evaluation | Artificial Analysis (opens in a new tab)independent |
| GPQADiamond / Artificial AnalysisAccuracy · Higher is better | 92.828% | GPT-5.6 Sol (High); 198-question Diamond subset; independent AA evaluation | Artificial Analysis (opens in a new tab)independent |
| GPQADiamond / Artificial AnalysisAccuracy · Higher is better | 78.99% | GPT-5.6 Sol (Non-reasoning); 198-question Diamond subset; independent AA evaluation | Artificial Analysis (opens in a new tab)independent |
| GPQADiamond / Artificial AnalysisAccuracy · Higher is better | 92.626% | GPT-5.6 Sol (Medium); 198-question Diamond subset; independent AA evaluation | Artificial Analysis (opens in a new tab)independent |
| GPQADiamond / Artificial AnalysisAccuracy · Higher is better | 89.798% | GPT-5.6 Sol (Low); 198-question Diamond subset; independent AA evaluation | Artificial Analysis (opens in a new tab)independent |
| Humanity's Last ExamHLE — source does not specify subsetAccuracy · Higher is better | 64.5% | temperature=1.0; top_p=0.95; max generation 163840 tokens; max context 300000 tokens; context management; GPT-5.6 Luna medium judge. · subset: unspecified · Tools enabled | Z.ai ↗ (opens in a new tab)other · 2026-08-14 |
| Humanity's Last ExamStable HLE — full setAccuracy · Higher is better | 58% | Full-set comparison; reasoning effort and tool configuration as reported. Kept separate from later publications. · subset: full · Tools enabled | Moonshot AI ↗ (opens in a new tab)provider reported |
| Humanity's Last ExamStable HLE — full setAccuracy · Higher is better | 44.5% | Full-set comparison; reasoning effort and tool configuration as reported. Kept separate from later publications. · subset: full · No tools | Moonshot AI ↗ (opens in a new tab)provider reported |
| Humanity's Last ExamStable HLE — text-only subsetAccuracy · Higher is better | 49.5% | question count: 2158 · reported precision decimals: 1 · GPT-5.6 Sol (max); original HLE; 2,158 text-only questions; no tools; pass@1; chart labels rounded to one decimal. · source model href: /models/gpt-5-6-sol · subset: text · No tools | Artificial Analysis (opens in a new tab)independent |
| MyPCBench184 tasks · 100 stepsPerfect-task rate · Higher is better | 55.4% | interface: computer + bash · judge: gemini-3.1-flash-lite-preview · max steps: 100 · rubric score percent: 85.3 · selection: Canonical matched-protocol rows only; four 200-step self-reported rows retained in source snapshot, excluded from this track. · tasks: 184 | MyPCBench / Carnegie Mellon University ↗ (opens in a new tab)benchmark publisher |
| Terminal-Bench4.0Task success · Higher is better | 37.3% | Codex; effort=max; 37.3% ± 3.8% | Terminal-Bench ↗ (opens in a new tab)benchmark publisher |
| Terminal-Bench4.0 / mini-swe-agentTask success · Higher is better | 39.899% | harness: mini-swe-agent · repeats: 3 · selection: best reported configuration per release · GPT-5.6 Sol (Max) | Artificial Analysis (opens in a new tab)independent |