OpenAI·released 2025-06-101 source
o3-pro benchmark scores: 8 benchmarks tracked, leading 1. 0% of its scores are independently reproduced — the rest are self-reported or unverified.
Leads on 1 of 8 benchmarks·best result 97.2 on Fiction.LiveBench (unverified)·0 of 8 independently reproduced·$20/$80 per M tokens
Head-to-heado3-pro vs GPT-5Consensus: LiteLLM · Cross-check: models.dev, OpenRouter · See every model’s pricing →
Leads on: Fiction.LiveBench
| Benchmark | Score | Evidence status | Measured |
|---|---|---|---|
| Fiction.LiveBench | 97.2accuracy (%) | unverifiedT1 | 2025-06-10Fiction.live leaderboard |
| Aider polyglot | 84.9accuracy (%) | unverified· optimizedT1 | 2025-06-10https://aider.chat/docs/leaderboards/ |
| Lech Mazur Writing | 84.4accuracy (%) | unverifiedT1 | 2025-06-10lechmazur/writing Github repository |
| DTBench | 78.22accuracy (%) | unverifiedT1 | 2025-06-10https://conceptualreasoning.ai/dtbench |
| ARC-AGI | 59.33accuracy (%) | unverified· optimizedT1 | 2025-06-10https://arcprize.org/leaderboard |
| WeirdML | 58.21accuracy (%) | unverifiedT1 | 2025-06-10https://htihle.github.io/weirdml.html |
| LMCA | 45.27accuracy (%) | unverifiedT1 | 2025-06-10https://conceptualreasoning.ai/lmca |
| ARC-AGI-2 | 4.86accuracy (%) | unverifiedT2 | 2025-06-10 |
Claims drawn from cited facts, not live model generation.
This model originates from OpenAI and dates to 2025, placing it in the current generation of releases from that developer. No parameter count is available here, so its size relative to compact, mid-size or frontier-scale systems cannot be stated.
3 cited facts
This model is scored on 8 tracked benchmarks and currently holds the top score on 1 of them. Its record is unverified: neither confirmed nor disputed by outside evaluation.
3 cited facts
On average across its benchmarks, this model trails the best reported score by 27.13 points, a wide gap that reads as well behind the leaders rather than at or near the frontier. It does hold the top score on 1 benchmark, so there is one narrow spot of category leadership, but that does not offset the overall average gap.
2 cited facts
6 facts cross-checked across data sources: 3 corroborated, 2 single-source, 1 disputed.
Source facts, citations, and refresh stamp for this record.
Sources: epoch_benchmarks, litellm_prices, modelsdev_models, openrouter_models
o3-pro is an AI model developed by OpenAI, released 2025-06-10. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.
o3-pro has recorded scores on 8 benchmarks, each shown with its evidence status.
o3-pro has recorded scores on 8 benchmarks — Lech Mazur Writing, DTBench, WeirdML, LMCA, Aider polyglot, ARC-AGI, and 2 more. The full table above shows each score with its evidence status.
0 of 8 recorded scores (0%) are independently reproduced rather than self-reported by the lab.
o3-pro has 8 tracked claims: 8 unverified.
Listed API pricing: $20 per million input tokens, $80 per million output tokens. See the pricing block for the full breakdown.