Gemma 3 12B
Google · 1 configuration
Intelligence3.8
Input / 1M$0
Output / 1M$0
Context128K
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 cost1 measured configurations · 0 without both values
Selected: thinking default · Click a point to switch configuration
Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities,...
Intelligence IndexArtificial Analysis composite index, not a percentage. Different versions of the index are not directly comparable.
3.8 pointsGPQA DiamondGraduate-level scientific questions. Measures specialist reasoning rather than web-search quality.
34.9 %Humanity’s Last ExamDifficult questions across many disciplines, compared under the same evaluation setup.
4.2 %SciCodeCode generation for scientific research problems.
16.4 %Terminal-Bench 2.1Agent tasks in a terminal environment. Results also depend on the evaluation harness.
0.0 %τ²-BenchTool use and interaction in assistance scenarios. Does not replace an evaluation of your integration.
10.8 %IFBenchVerifiable instruction following. A proxy for precision rather than a benchmark of creative style.
36.7 %MMMU-ProMultimodal understanding of academic problems that include images.
37.5 %Omniscience accuracyFactual accuracy in Artificial Analysis's Omniscience evaluation.
10.3 %Specifications from the catalog
Inputtext, image
Outputtext
Tool callingYes
Structured outputYes
SpeedNot available
First tokenNot available
Maximum output16.4K
Catalog listing date2025-03-13