o1-preview: benchmark scores, price and ranking
o1-preview from OpenAI ranks #271 of 743 LLMs on the LLMs Tiger aggregate benchmark score (49.4), based on 10 benchmark results captured up to 2026-10-13.
- Organisation
- OpenAI
- Type
- Preview release
- Open weights
- No
- Cheapest API price
- $15.00 in / $60.00 out per million tokens (via azure, observed 2026-10-13T19:47Z)
- Aggregate rank
- #271 of 743 (score 49.4)
Relative to other models, o1-preview is strongest on Aider Code Editing (#3 of 92) and weakest on Cybench (#12 of 16).
Across the 10 benchmarks where at least ten models are scored, o1-preview is in the top 10% on 1 and in the top half on 6.
Its cheapest listed input price, $15.00 per million tokens, is lower than that of 2% of the 427 models with a listed price.
o1-preview benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| Aider Code Editing | 79.7% | #3 of 92 | Aider Aider Code Editing leaderboard | 2026-10-13 |
| Aider Refactoring | 75.3% | #2 of 12 | Aider Aider Refactoring leaderboard | 2026-10-13 |
| WeirdML | 47.6% | #33 of 109 | Epoch AI Benchmarking Hub | 2026-10-13 |
| Epoch ECI | 135 | #73 of 204 | Epoch AI Benchmarking Hub | 2026-10-13 |
| SWE-bench Verified | 64.6% | #27 of 73 | SWE-bench Verified leaderboard | 2026-10-13 |
| SimpleBench | 41.7% | #30 of 61 | Epoch AI Benchmarking Hub | 2026-10-13 |
| METR time horizon | 20 min | #18 of 31 | Epoch AI Benchmarking Hub | 2026-10-13 |
| OTIS Mock AIME | 31.1% | #120 of 189 | Epoch AI Benchmarking Hub | 2026-10-13 |
| GPQA Diamond | 50.3% | #186 of 275 | Epoch AI Benchmarking Hub | 2026-10-13 |
| Cybench | 10.0% | #12 of 16 | Epoch AI Benchmarking Hub | 2026-10-13 |
About o1-preview
o1-preview is a preview release from OpenAI. The cheapest API listing we track charges $15.00 per million input tokens and $60.00 per million output tokens, through azure. We hold 10 benchmark results for it across 3 areas, and 1 is in the top tenth of its board.
Where o1-preview stands
On our aggregate score, o1-preview is #271 of 743 ranked models, around the middle of the table. Its score is 49.4; the scale is centred on 50, so it sits 0.6 points below the typical model. That beats 64% of the ranked field. Among the 88 ranked OpenAI models it is #38. The next one up is GPT-4 (0125) at 49.4. The next one down is gpt-3.5-turbo-1106 at 49.2.
o1-preview, benchmark by benchmark
On Aider Code Editing, o1-preview scores 79.7% and ranks #3 of 92. Only 2 models score higher, led by Claude 3.5-Sonnet-20241022 at 84.2%, which is 4.5 points ahead. It covers Aider official benchmark. The board median is 55.6%, so o1-preview is 24.1 points above it. Source: Aider Aider Code Editing leaderboard, captured 2026-10-13.
On Aider Refactoring, o1-preview reaches 75.3%. It ranks #2 of 12, ahead of 83% of the field, and 16.8 points behind Claude 3.5-Sonnet-20241022 at 92.1%. The results either side are Claude 3.5-Sonnet-20241022 (92.1%) and claude-3-opus-20240229 (72.3%). It covers Aider official benchmark. The board median is 50.0%, so o1-preview is 25.3 points above it. Source: Aider Aider Refactoring leaderboard, captured 2026-10-13.
o1-preview manages 47.6% on WeirdML, good for #33 of 109. GPT-6 Astra (max) leads at 93.3%, 45.7 points further up, and 70% of models sit below. The results either side are gpt-oss-120b (high) (48.2%) and DeepSeek-V3.2-Speciale (46.7%). It covers unusual machine-learning tasks. The board median is 41.2%, so o1-preview is 6.4 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Epoch ECI puts o1-preview at 135, #73 of 204. That beats 64% of the models tested, but 22 points short of Qwen 3.8 Max (156). The results either side are GPT-4.1 mini (135) and Gemini 2.0 Flash (Dec 2024) (135). It covers Epoch Capabilities Index: Epoch's own aggregate of ~40 benchmarks. The board median is 127, so o1-preview is 8 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
On SWE-bench Verified, o1-preview reaches 64.6%. It ranks #27 of 73, ahead of 63% of the field, and 18.9 points behind Claude Opus 4.7 (max) at 83.5%. The results either side are Kimi K2 (65.4%) and Kimi K2 Thinking (63.4%). It covers real GitHub issues resolved. The board median is 56.4%, so o1-preview is 8.2 points above it. Source: SWE-bench Verified leaderboard, captured 2026-10-13.
o1-preview manages 41.7% on SimpleBench, good for #30 of 61. Claude Fable 5 (max) leads at 81.9%, 40.2 points further up, and 51% of models sit below. The results either side are claude-3-7-sonnet-20250219 (44.9%) and Claude 3.5 Sonnet (Oct 2024) (41.4%). It covers trick questions that test common sense. The board median is 41.4%, so o1-preview is 0.3 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
o1-preview lands below the middle on METR time horizon: 20 min, #18 of 31, with 55% of the field scoring higher. Claude Mythos Preview (Early) leads at 1,045 min; the closest result above is Claude 3.5 Sonnet (Oct 2024) at 21 min. It covers length of task (minutes) done at 50% success. The board median is 23 min, so o1-preview is 3 minutes below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
o1-preview lands below the middle on OTIS Mock AIME: 31.1%, #120 of 189, with 63% of the field scoring higher. Claude Fable 5.1 (max) leads at 100.0%; the closest result above is Gemini 2.0 Flash (Feb 2025) at 31.1%. It covers competition maths (AIME 2024-25 style). The board median is 57.8%, so o1-preview is 26.7 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
o1-preview lands below the middle on GPQA Diamond: 50.3%, #186 of 275, with 67% of the field scoring higher. GPT-6 Astra (max) leads at 95.8%; the closest result above is c4ai-command-a-03-2025 at 50.5%. It covers expert-level science questions. The board median is 71.7%, so o1-preview is 21.4 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
o1-preview lands below the middle on Cybench: 10.0%, #12 of 16, with 69% of the field scoring higher. grok-4-0709 leads at 43.0%; the closest result above is claude-3-opus-20240229 at 10.0%. It covers cybersecurity capture-the-flag tasks. The board median is 17.5%, so o1-preview is 7.5 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Strengths and weaknesses
Software engineering: strong, ahead of 80% of the field on average across 2 benchmarks. Its best board here is Aider Code Editing (#3 of 92) and its weakest is SWE-bench Verified (#27 of 73). Reasoning, maths and science: uneven, ahead of 40% of the field on average across 3 benchmarks. Its best board here is SimpleBench (#30 of 61) and its weakest is GPQA Diamond (#186 of 275). General knowledge and chat: solid, ahead of 57% of the field on average across 5 benchmarks. Its best board here is Aider Refactoring (#2 of 12) and its weakest is Cybench (#12 of 16).
Price and value
The median input price across the 427 priced models is $0.500 per million tokens, so o1-preview costs more than the typical listing. Models that score within 1.5 points or better and cost less: Qwen2.5-Coder-7B-Instruct ($0.010, 49.5), DeepSeek V4.1 Flash (high) ($0.010, 48.9), Qwen3.7 Flash ($0.030, 50.5). Similar scores cost more on GPT-4 (0613) ($30.00) and GPT-4 (0125) ($30.00).
Reading these numbers
The 10 scores on this page come from 4 sources: Epoch AI Benchmarking Hub (7), Aider Aider Code Editing leaderboard (1), Aider Aider Refactoring leaderboard (1), SWE-bench Verified leaderboard (1). 2 of them come from an official leaderboard and 8 from datasets or third-party runs.
Models ranked nearby
- #268 Grok 3 mini (49.4)
- #269 GPT-4 (0125) (49.4)
- #270 Gemini 3 Flash Preview (High) (49.4)
- #272 gemini/gemini-1.5-pro-latest (49.4)
- #273 Gemini 2.0 Pro exp-02-05 (49.4)
- #274 Qwen3-Next-80B-A3B-Instruct (49.3)
More from OpenAI
- GPT-6 Astra (max) (#2, 64.0)
- GPT-6.1 Sol (max) (#5, 60.6)
- GPT-5.6 Sol (max) (#6, 60.1)
- GPT-6 Sol (max) (#7, 60.0)
- GPT-5.5 (xhigh) (#11, 58.0)
- GPT-5.6 Terra (max) (#12, 57.8)
Compare it on the live leaderboard → · How the aggregate score works