ModelBeat
Models/Gemma 4 31B IT
All models
G

Google DeepMind: Gemma 4 31B IT

Watch this modelCompareFind source
google/gemma-4-31b-it

Gemma 4 31B IT is Google DeepMind’s open-weights AI model, released Apr 2, 2026. It costs $0.13/M input and $0.38/M output tokens, and handles a 262K-token context window.

Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output.

Context
262K
Input
$0.13 / 1M
Output
$0.38 / 1M
Released
Apr 2, 2026
Modalities
imagetextvideo→text

Benchmarks

Scores on standardized evaluations. Higher is better — and rank shows where Gemma 4 31B IT lands among all models tracked on Model Beat.

19.0
Intelligence Index
Epoch AI
19th percentile of tracked models
25.0
Coding Index
Epoch AI
25th percentile of tracked models

Reasoning

3 evals
GPQA Diamond75.8%

Graduate-level scientific reasoning

SimpleQA Verified10.4%

Factual accuracy & hallucination

WeirdML52.3%

Novel ML problem-solving

Coding

1 evals
WebDev Arena1364

Human-rated web development

Math

1 evals
AIME 2024/202573.3%

Olympiad-qualifier math

Changelog

8
  • Gemma 4 31B IT: cheapest credible provider via DeepInfra now $0.38 per 1M output tokens (was $0.35, +9%)cheapest provider · Sep 20, 2026
  • Gemma 4 31B IT: cheapest credible provider via DeepInfra now $0.13 per 1M input tokens (was $0.12, +8%)cheapest provider · Sep 20, 2026
  • Gemma 4 31B IT: reported SimpleQA Verified score rose from 9.6% to 10.4% (+9%)benchmark · Aug 28, 2026
  • Gemma 4 31B IT: cheapest credible provider via Venice now $0.12 per 1M input tokens (was $0.08, +50%)cheapest provider · Aug 26, 2026
  • Gemma 4 31B IT: cheapest credible provider via OpenInference now $0.08 per 1M input tokens (was $0.1, -20%)cheapest provider · Aug 17, 2026
  • Gemma 4 31B IT: cheapest credible provider via CoreWeave now $0.1 per 1M input tokens (was $0.08, +25%)cheapest provider · Aug 14, 2026
  • Gemma 4 31B IT: cheapest credible provider via OpenInference now $0.08 per 1M input tokens (was $0.1, -20%)cheapest provider · Aug 10, 2026
  • Gemma 4 31B IT: cheapest credible provider via OpenInference now $0.1 per 1M input tokens (was $0.12, -17%)cheapest provider · Jul 22, 2026

Compare Gemma 4 31B IT with

Gemma 4 31B IT vs GPT-6 AstraGemma 4 31B IT vs Claude Fable 5Gemma 4 31B IT vs Kimi K3Gemma 4 31B IT vs Grok 4.6Gemma 4 31B IT vs DeepSeek-V4-Pro-0813

Frequently asked questions

What is Gemma 4 31B IT?

Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output.

Who created Gemma 4 31B IT?

Gemma 4 31B IT is an AI model developed by Google DeepMind, part of the Gemma family, released Apr 2, 2026.

How much does Gemma 4 31B IT cost?

Gemma 4 31B IT is served from $0.13 per million input tokens and $0.38 per million output tokens (cheapest credible provider on OpenRouter: DeepInfra).

What is Gemma 4 31B IT's context window?

Gemma 4 31B IT supports a context window of up to 262K tokens.

How does Gemma 4 31B IT perform on benchmarks?

On standardized evaluations tracked by Epoch AI, Gemma 4 31B IT scores 52.3% on WeirdML, 1364 on WebDev Arena, 73.3% on AIME 2024/2025.

Is Gemma 4 31B IT open source?

Gemma 4 31B IT is released with open weights (Open weights (restricted use)), so the model can be downloaded and self-hosted.

Benchmark data from Epoch AI (CC BY) and Artificial Analysis; pricing, specs & descriptions from OpenRouter. Data updated Sep 20, 2026.