Gemma 4 31B
Google · 2 thinking configurations
Intelligence13.9
Input / 1M$0.14
Output / 1M$0.40
Context256K
Tariff: Google · Source: artificial-analysis · USD per million tokens
Thinking configuration
What does more thinking buy you?
Compare this model’s measured thinking configurations.
Scroll the chart ↔ · tap a point for details
Efficient frontier: best measured score at each cost2 measured configurations · 0 without both values
Selected: thinking none · Click a point to switch configuration
Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output. Features a 256K token context window, configurable thinking/reasoning mode, native function...
Intelligence IndexArtificial Analysis composite index, not a percentage. Different versions of the index are not directly comparable.
13.9 pointsGPQA DiamondGraduate-level scientific questions. Measures specialist reasoning rather than web-search quality.
76.3 %Humanity’s Last ExamDifficult questions across many disciplines, compared under the same evaluation setup.
11.8 %Terminal-Bench 2.1Agent tasks in a terminal environment. Results also depend on the evaluation harness.
29.2 %τ²-BenchTool use and interaction in assistance scenarios. Does not replace an evaluation of your integration.
65.5 %IFBenchVerifiable instruction following. A proxy for precision rather than a benchmark of creative style.
53.5 %MMMU-ProMultimodal understanding of academic problems that include images.
70.3 %Omniscience accuracyFactual accuracy in Artificial Analysis's Omniscience evaluation.
16.6 %Specifications from the catalog
Inputimage, text, video
Outputtext
Tool callingYes
Structured outputYes
Speed39.4 tok/s
First token2.04 s
Maximum output16.4K
Catalog listing date2026-04-02