OpenAI
GPT-5.5 benchmark results
18 results on 8 benchmarks from 3 sources.
Also listed as: gpt-5.5-high · gpt-5.5 · gpt-5.5_none · GPT-5.5 (no thinking) · gpt-5.5_low · GPT-5.5 (low) · gpt-5.5-pre-release_xhigh · GPT-5.5 (xhigh) · gpt-5.5_xhigh · gpt-5.5_unknown
All benchmark results
| Benchmarks | Score | Rank | Setting | Source type | Date |
|---|---|---|---|---|---|
| LMArena Text | 1482 | 24 | Overall, style control on; 69437 votes; 95% CI 1478–1485“gpt-5.5-high” | Official leaderboard | Oct 8, 2026 |
| LMArena Text | 1477 | 29 | Overall, style control on; 71047 votes; 95% CI 1473–1480“gpt-5.5” | Official leaderboard | Oct 8, 2026 |
| LMArena Vision | 1287 | 16 | Overall, style control on; 27223 votes; 95% CI 1281–1292“gpt-5.5” | Official leaderboard | Oct 2, 2026 |
| LMArena Vision | 1281 | 24 | Overall, style control on; 25488 votes; 95% CI 1275–1287“gpt-5.5-high” | Official leaderboard | Oct 2, 2026 |
| GPQA Diamond | 94.0% | 10 | Epoch AI run, accuracy; effort: xhigh | Independent run | Apr 24, 2026 |
| GPQA Diamond | 90.7% | 35 | Epoch AI run, accuracy; effort: low | Independent run | May 5, 2026 |
| GPQA Diamond | 77.3% | 126 | Epoch AI run, accuracy; effort: no thinking | Independent run | Aug 7, 2026 |
| SWE-bench Verified | 80.6% | 2 | Epoch AI run (Epoch scaffold); effort: xhigh | Independent run | Apr 24, 2026 |
| AIME (OTIS Mock AIME 2024–2025) | 100.0% | 1 | Epoch AI run, OTIS Mock AIME 2024–2025; effort: xhigh | Independent run | Apr 24, 2026 |
| AIME (OTIS Mock AIME 2024–2025) | 84.4% | 95 | Epoch AI run, OTIS Mock AIME 2024–2025; effort: low | Independent run | Aug 7, 2026 |
| AIME (OTIS Mock AIME 2024–2025) | 57.8% | 152 | Epoch AI run, OTIS Mock AIME 2024–2025; effort: no thinking | Independent run | Aug 7, 2026 |
| OSWorld 2.0 | 13.0% | 9 | Binary accuracy; 500 steps; reasoning: xhigh; effort: xhigh | Official leaderboard via Epoch AI | Apr 23, 2026 |
| τ²-bench | 44.6% | 5 | banking_knowledge domain, pass^1; effort: xhigh; submitted by Sierra; v1.0.1 | Official leaderboard | Jul 22, 2026 |
| Terminal-Bench 2.0 | 84.7% | 1 | Terminal-Bench 2.0; agent: NexAU-AHE; effort: unknown thinking | Official leaderboard via Epoch AI | Apr 23, 2026 |
| Terminal-Bench 2.0 | 83.2% | 2 | Terminal-Bench 2.0; agent: Capy; effort: unknown thinking | Official leaderboard via Epoch AI | Apr 23, 2026 |
| Terminal-Bench 2.0 | 82.3% | 3 | Terminal-Bench 2.0; agent: Codex CLI; effort: unknown thinking | Official leaderboard via Epoch AI | Apr 23, 2026 |
| Terminal-Bench 2.0 | 82.0% | 4 | Terminal-Bench 2.0; agent: Codex; effort: unknown thinking | Official leaderboard via Epoch AI | Apr 23, 2026 |
| Terminal-Bench 2.0 | 66.1% | 31 | Terminal-Bench 2.0; agent: clnkr; effort: unknown thinking | Official leaderboard via Epoch AI | Apr 23, 2026 |
FAQ
Where do these GPT-5.5 scores come from?
From Epoch AI Benchmarking Hub, LMArena leaderboard dataset, and τ²-bench leaderboard submissions. Each row links to the original leaderboard, dataset or report, and vendor-reported numbers are labelled.