Back to Models

Olmo 3.1 32b Think

A large-scale, 32-billion-parameter model designed for deep reasoning, complex multi-step logic, and advanced instruction following.

Thinking Mode
Parameters
32000000000 B
Context
65,536 tokens
Released
Invalid Date

Leaderboards

Average Score combining domain-specific Autobench scores; Higher is better

Performance vs. Industry Average

Context Window

Olmo 3.1 32b Think has a smaller context window than average (406k tokens), with a context window of 66k tokens.

Olmo 3.1 32b Think - AutoBench