OpenAI
availableShows if the model has enough results for an index.GPT-5
GPT-5 is a model from OpenAI. 24 benchmarks count toward its score, in 6 categories.
IndexOverall score. 50 is the middle.56.5 ±4.3
CoverageShare of the index weight with results.90%
SpeedOutput tokens per second.62/s
Input / 1MUS dollars per 1M input tokens.$1.25 batch $0.625
Output / 1MUS dollars per 1M output tokens.$10 batch $5 US dollars per 1M output tokens in a batch.
ContextMaximum tokens in one request.400K
EloLMArena rating and rank.1406 (#136)
50 is the middle of the board. The range shows the doubt in the index. Batch work costs less.
31,428 votes. Elo shows what people prefer. It does not change the score.
CapabilitiesScore per category. 50 is the middle.
50 is the middleResults
24 counted| BenchmarkThe test name. | CategoryThe capability that the test measures. | ResultThe score from the publisher. | IndexThis result on the index scale. | RunThe settings of the run. | DateDate of the result. | Published byThe source of the result. |
|---|---|---|---|---|---|---|
| MATH level 5 | Math | 98.1% | 46.9 | high effort | — | Epoch AI |
| MATH 500 | Math | 96.0% | 49.1 | — | 9 Jan 2026 | Vals AI |
| τ²-bench Telecom | Agentic | 95.8% | 67.6 | Sierra | 2 Mar 2026 | Sierra Research |
| AIME | Math | 93.4% | 57.3 | — | 16 Apr 2026 | Vals AI |
| MGSM | Multilingual | 92.8% | — | — | 9 Jan 2026 | Vals AI |
| OTIS Mock AIME 2024-2025 | Math | 91.4% | 62.6 | high effort | — | Epoch AI |
| MMLU Pro | Knowledge | 86.5% | 56.9 | — | 1 Sept 2026 | Vals AI |
| GPQA diamond | Knowledge | 86.2% | 57.8 | high effort | — | Epoch AI |
| LiveCodeBench | Coding | 85.9% | 62.4 | — | 1 Sept 2026 | Vals AI |
| GPQA Diamond | Knowledge | 85.6% | 57.3 | — | 1 Sept 2026 | Vals AI |
| τ²-bench Retail | Agentic | 81.6% | 57.4 | Sierra | 30 Apr 2026 | Sierra Research |
| MMMU Pro | Multimodal | 81.5% | 60.0 | — | 1 Sept 2026 | Vals AI |
| SWE-Bench verified | Coding | 73.6% | 56.5 | high effort | — | Epoch AI |
| SWE-bench | Coding | 69.0% | 52.9 | — | 1 Sept 2026 | Vals AI |
| ARC-AGI-1 (semi-private) | Reasoning | 65.7% | 58.2 | high effort | — | ARC Prize Foundation |
| SWE-bench Verified | Coding | 65.0% | 49.7 | medium effort · mini-SWE-agent | 26 Feb 2026 | SWE-bench team |
| SWE-bench Verified | Coding | 65.0% | 49.7 | medium effort · mini-SWE-agent | 1 Sept 2026 | SWE-bench team |
| τ²-bench Airline | Agentic | 62.5% | 43.7 | Sierra | 2 Mar 2026 | Sierra Research |
| FrontierMath-Tiers-1-3-v2-Private | Math | 55.4% | 62.6 | high effort | — | Epoch AI |
| SimpleQA Verified | Knowledge | 50.1% | 67.9 | high effort | — | Epoch AI |
| Terminal-Bench 1.0 | Agentic | 48.8% | 52.9 | — | 12 Jan 2026 | Vals AI |
| Terminal-Bench 2.0 | Agentic | 37.1% | 49.1 | — | 4 Jun 2026 | Vals AI |
| Chess Puzzles | Reasoning | 37.0% | 71.4 | high effort | — | Epoch AI |
| FrontierMath-2025-02-28-Private | Math | 32.4% | 60.9 | high effort | — | Epoch AI |
| Mystery Game Puzzles | Reasoning | 23.0% | 58.4 | high effort | — | Epoch AI |
| FrontierMath-Tier-4-v2-Private | Math | 22.0% | 58.5 | high effort | — | Epoch AI |
| Vibe Code Bench v1.1 | Coding | 20.1% | 50.5 | OpenHands | 21 Sept 2026 | Vals AI |
| IOI v1 | Coding | 20.0% | 51.3 | — | 9 Aug 2026 | Vals AI |
| EBR-bench | Reasoning | 12.7% | 58.0 | high effort | — | Epoch AI |
| FrontierMath-Tier-4-2025-07-01-Private | Math | 12.5% | 57.0 | high effort | — | Epoch AI |
| ARC-AGI-2 (semi-private) | Reasoning | 9.9% | 44.7 | high effort | — | ARC Prize Foundation |
24 benchmarks count, from 30 of 31 results. A grey row does not count. Too few models took that benchmark.