o1
Reasoning
The first o-series reasoning model, proved chain-of-thought scaling worked. Now superseded by o3 / o4-mini for new traffic.
o1 is a reasoning AI model from OpenAI. It costs $15.000 per million input tokens and $60.000 per million output tokens (blended $28.500/M), with a 200,000-token context window.
INPUT
$15.000/M
per million input tokens
OUTPUT
$60.000/M
per million output tokens
CONTEXT
200,000
tokens
What it is good at
- Strong math & code
- Multi-step reasoning
Typical use cases
- Legacy o1 reasoning pipelines
- Migration to o3
Benchmarks
vs. best public score
Hand-curated from each provider's published reports and public leaderboards. Methodology varies across sources, treat as directional rather than authoritative.
More from OpenAI
See all 28 →GPT-5.6 Sol
Frontier · 1,050,000 ctx
in $5.000/Mout $30.000/M
GPT-5.6 Terra
Balanced · 1,050,000 ctx
in $2.000/Mout $12.000/M
GPT-5.6 Luna
Efficient · 1,050,000 ctx
in $0.200/Mout $1.200/M
GPT-5.5
Frontier · 1,050,000 ctx
in $5.000/Mout $30.000/M
GPT-5.2
Frontier · 400,000 ctx
in $1.750/Mout $14.000/M
GPT-5.2-Codex
Coding · 400,000 ctx
in $1.750/Mout $14.000/M
Frequently asked questions
How much does o1 cost?
o1 costs $15.000 per million input tokens and $60.000 per million output tokens, for a blended reference rate of $28.500 per million tokens.
What is o1's context window?
o1 supports up to 200,000 tokens of context in a single request.
What is o1 best for?
o1 is well suited to Strong math & code, Multi-step reasoning and Legacy o1 reasoning pipelines.
Who makes o1?
o1 is developed and served by OpenAI. It was released in Sep 2024.