Advanced models with chain-of-thought and complex reasoning capabilities.
What to look for when choosing a model for reasoning
Chain-of-thought reasoning support
Large context window for complex problems
High accuracy on benchmarks
Ability to show working and explain steps
Top picks for reasoning based on our analysis
openai
Input
$15/1M
Output
$60/1M
Context
200K
Max Output
100K
openai
Input
$1.1/1M
Output
$4.4/1M
Context
200K
Max Output
100K
Input
$1.25/1M
Output
$10/1M
Context
1.0M
Max Output
66K
openai
Input
$2.5/1M
Output
$10/1M
Context
128K
Max Output
16K
openai
Input
$0.15/1M
Output
$0.60/1M
Context
128K
Max Output
16K
meta
Input
$0.20/1M
Output
$0.60/1M
Context
1.0M
Max Output
16K
Dedicated reasoning models (o1, o3) excel at complex tasks
Allow sufficient output tokens for reasoning traces
Consider the trade-off between speed and accuracy