Gemma 4 31B IT
Largest Gemma 4 instruction model for open, self-hosted chat and reasoning
- Strongest at
- Math Top 44% of the past year's models · one benchmark so far
- Price per 1M tokens
- $0.2 / $0.4 input / output · OpenRouter
- Context window
- 262.1K tokens in · 32.8K out
- Since the last one
- 386 days after Gemma 3 4B IT
Capability profile
Scored against 91 models released in the past year. 50 is the average model; further out is better.
No results yet: Coding, Reading & documents, Reasoning, Visual reasoning, Computer use
Where these scores come from
Each capability combines the published benchmarks listed here, rescaled so 50 is the reference average and 15 points is one standard deviation. A hollow dot rests on a single benchmark.
- Writing EQ-Bench Creative Writing v3, EQ-Bench Longform Writing
- Math OTIS Mock AIME 2024-2025
- Science GPQA diamond
Price against capability
The shaded corner holds anything better and cheaper. Prices blend input and output 3 to 1.
3 models from the past year score higher on Math for less: Claude Haiku 5.5, GPT-6 Luna, DeepSeek V4.1 Flash.
1 model from the past year scores higher on Writing for less: DeepSeek V4.1 Flash.
8 models from the past year score higher on Science for less: GPT-6 Luna, GLM-5.3-Flash, Claude Haiku 5.5 and 5 more.
OpenRouter $0.2 / $0.4
Since Gemma 3 4B IT
6 releases in the Gemma line.
A newer release is out: Gemma 4 12B IT.
Epoch benchmark boards
Best result on each board it has run. The tick marks the leader.
Google's past year
37 release days in the past year, about one every 10 days. The next is expected around Nov 3, 2026. Forecasts ↗
Specifications
Specifications and prices from Models.dev; availability and limits may differ by provider. Benchmarks from Epoch AI (CC BY 4.0), LiveBench and the other sources named above.