o1: Benchmarks, Pricing & Context Window
o1 is a language model from OpenAI, released in December 2024.
A research preview model focused on mathematical and logical reasoning capabilities, demonstrating improved performance on tasks requiring step-by-step reasoning, mathematical problem-solving, and code generation. The model shows enhanced
o1 benchmarks
Capability tiers
Standing within each category, adjusted for leaderboard depth.
Real tasks performance
High-confidence performance for o1 across real-world prompt categories. Only 95% intervals at most 4 points wide are shown.
Performance by conversation depth
How o1 holds up as conversations get longer.
Quality Tracker
o1 Performance Across Datasets
Scores sourced from the model's scorecard, paper, or official blog posts
llm-stats.com - Sat Sep 12 2026o1 pricing
Providers
o1 starts at $15.00 per million input tokens and $60.00 per million output tokens via Azure. See all 2 providers below with their per-token pricing, latency, throughput, and modality support.
| Provider | Input $/M | Cached input $/M | Output $/M | Context in / out | TTFT p95 s | Output p5 c/s | Modalities in / out |
|---|---|---|---|---|---|---|---|
 Azure | $15.00 | — | $60.00 | 200.0K/100.0K | 0.54 | — | / |
 OpenAI | $15.00 | — | $60.00 | 200.0K/100.0K | 16.20 | — | / |
Cached input is the discounted price for prompt tokens served from a provider cache. TTFT is time to first token. Output is characters per second; p5 is the sustained floor exceeded by 95% of observed requests.
o1 context window
Input and output token limits for o1, plus how it ranks on long-context understanding.
Try now
huggleMake it with
o1.
o1 latency
o1 time to first token, sustained output throughput, and failed-request rate from live model usage over the trailing 7 days.
o1 examples
Recent arena outputs from o1, picked from the highest-ranked matchups.
o1 license
o1 is a proprietary model available under its provider's product and API terms.
- License
- Proprietary
- Hosted access
Proprietary license - usage restrictions apply
o1 resources
Official sources for o1: provider documentation, paper or system card, official launch post, source repository.
o1 vs other models
The most-compared alternatives to o1 are Mistral Large 3, Sarvam-105B, Qwen3-235B-A22B-Instruct-2507. Open any pair side-by-side for benchmarks, pricing, context, and latency.
Models like o1
Models ranked just above and below o1 by LLM Stats score.
FAQ
Common questions about o1.