All AI models · LLM benchmarks · Methodology

Google

Gemma 4 31B

Largest Gemma 4 instruction model for open, self-hosted chat and reasoning

Model facts

Context window
262K tokens
Maximum output
33K tokens
Cheapest paid input
$0.102 per 1M tokens
Cheapest paid output
$0.297 per 1M tokens
Open weights
yes
Providers
42

Capabilities and modalities

reasoning, tool calling, structured output, attachments, temperature control, open weights, text, image, audio, video, pdf.

Local hardware estimate

At 4K context: Q4_K_M 20.1 GiB, Q8_0 34.7 GiB, F16 62.0 GiB working memory.

Parameters
31.3 billion
Architecture
gemma4
Model license
apache-2.0

Planning estimate adapted from the MIT-licensed whichllm estimator using metadata from Hugging Face; not a vendor minimum requirement.

Artificial Analysis benchmarks

Compare this model on the full LLM benchmark leaderboard.

ConfigurationAAICodingMathOutput tokens/s
Reasoning29.743.4not measured35.2
Non-reasoning22.333.2not measured73.0

Observed price history

  • 2026-08-24: $0.102 input / $0.297 output per 1M tokens

API providers