DataLearner logo
GE

Gemma 4 31B

Chat modelGemmaGemma 4

Google Gemma 4 31B

Release date: 2026-04-02Updated: 2026-07-17 23:18:09.734Knowledge cutoff: 2025-011,785
Live demoGitHubHugging FaceCompare
Parameters
30.7B
Context length
256K
Chinese support
Not supported
Reasoning ability

Gemma 4 31B is a chat model from DeepMind, released on 2026-04-02. It accepts text and image input and returns text output. The cataloged parameter count is 30.7B. The recorded context window is 256K, and the recorded maximum output is 32K. Cataloged capabilities include Reasoning model. The checkpoint is listed under the Apache 2.0 license, with the license link included in the model references. Use the linked references to confirm current access, licensing, and provider-specific limits.

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

Gemma 4 31B

Model basics

Reasoning traces
Supported
Thinking modes
Thinking Mode (Default)Standard Mode
Context length
256K tokens
Max output length
32K tokens
Model type
Chat model
Modality (in / out)
Text, Image → Text
Release date
2026-04-02
Model file size
No data
MoE architecture
No
Total params / Active params
30.7B / N/A
Knowledge cutoff
2025-01
Gemma 4 31B

Open source & experience

Code license
Apache 2.0
Weights license
Apache 2.0- Commercial use permitted
GitHub repo
GitHub link unavailable
Hugging Face
https://huggingface.co/google/gemma-4-31b
Live demo
https://ai.google.dev/gemma
Gemma 4 31B

Official resources

Paper
Paper
DataLearnerAI blog
DataLearnerAI blog
Gemma 4 31B

API details

API speed
No data
No public API pricing yet.
Gemma 4 31B

Benchmark Results

Gemma 4 31B currently shows benchmark results led by MMLU Pro (24 / 132, score 85.20), LiveCodeBench (30 / 123, score 80), GPQA Diamond (58 / 188, score 84.30). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.

Thinking
Tool usage
Internet

General Knowledge

5 evaluations
Benchmark / mode
Score
Rank/total
MMLU Pro
Thinking Mode
85.20
24 / 132
GPQA Diamond
Thinking Mode
84.30
58 / 188
LiveBench
Standard Mode
61.62
62 / 115
HLE
Thinking Mode
19.50
121 / 173
HLE
Thinking ModeToolsInternet
26.50
97 / 173

Coding and Software Engineer

1 evaluations
Benchmark / mode
Score
Rank/total
LiveCodeBench
Thinking Mode
80
30 / 123

Agent Level Benchmark

1 evaluations
Benchmark / mode
Score
Rank/total
τ²-Bench
Thinking ModeTools
76.90
20 / 43

Math and Reasoning

1 evaluations
Benchmark / mode
Score
Rank/total
AIME 2026
Thinking Mode
89.20
16 / 19
View benchmark analysisCompare with other models

Compare with other models

Gemma 4 31B

Publisher

Google Gemma 4 31B

Model Overview

Gemma 4 31B is a chat model from DeepMind, released on 2026-04-02.

It accepts text and image input and produces text output. Its cataloged capabilities include Reasoning model. The cataloged parameter count is 30.7B. The recorded context window is 256K, and the recorded maximum output is 32K.

The checkpoint is listed under the Apache 2.0 license, with the license link included in the model references. The evaluation section contains 8 cataloged benchmark results with their recorded modes and scores. The page links 4 release, model-card, repository, or provider references for checking the underlying claims. Specifications, availability, and prices can change; undisclosed values are intentionally left unstated.

Gemma 4 31B

FAQ

What is Gemma 4 31B?

Gemma 4 31B is a chat model from DeepMind, released on 2026-04-02. It accepts text and image input and returns text output. The cataloged parameter count is 30.7B. The recorded context window is 256K, and the recorded maximum output is 32K. Cataloged capabilities include Reasoning model. The checkpoint is listed under the Apache 2.0 license, with the license link included in the model references. Use the linked references to confirm current access, licensing, and provider-specific limits.

What input and output modalities does Gemma 4 31B support?

The current model record lists text and image as input and text as output.

What are the main recorded specifications for Gemma 4 31B?

The cataloged parameter count is 30.7B. The recorded context window is 256K, and the recorded maximum output is 32K. Fields without a source-backed value remain undisclosed.

Why is no API price shown for Gemma 4 31B?

Open-weight checkpoint; Google publishes no billable first-party token price for this exact checkpoint.

Are benchmark results available for Gemma 4 31B?

The evaluation section contains 8 cataloged benchmark results with their recorded modes and scores. Compare only results that use the same benchmark version and evaluation mode.

Is Gemma 4 31B open source?

The checkpoint is listed under the Apache 2.0 license, with the license link included in the model references. Review the linked license text before commercial or derivative use.

DataLearner on WeChat

Follow DataLearner on WeChat for AI model updates and research notes.

DataLearner WeChat QR code