Google: Gemma 4 31B (free)
Google's Gemma 4 31B (free) accepts text, image, and video input and returns text, with a context window of 262,144 tokens and a maximum completion length of 32,768 tokens. It supports tool calling and a reasoning mode. The listing does not confirm structured output support, so verify that separately if your pipeline depends on strict schemas. At zero cost for both input and output tokens, this model suits budget-limited projects, prototyping, and high-volume experimentation. Its blended score of 53.7 across 4 benchmarks puts it in the middle of the field, and the breakdown is uneven: 71.7 on the coding index is its strongest result, while 23.8 on the agentic index suggests weaker performance on multi-step autonomous tasks. It scores 48.4 on the general AI index and 57.5 on EQ-Bench. Shortlist it for coding help or long-document and video analysis where price matters most, and test carefully before relying on it for agent workflows.
Benchmark results
Independent, published benchmarks. Blended score 53.7 across 4 benchmarks, last refreshed 2026-11-08. How scoring works →
| Benchmark | Measures | Score |
|---|---|---|
| AI Index | broad capability composite | 48.4 |
| AI Index Coding | software engineering tasks | 71.7 |
| AI Index Agentic | multi-step tool-using tasks | 23.8 |
| EQ-Bench | emotional understanding in dialogue | 57.5 |
- Model ID
- google/gemma-4-31b-it:free
- Vendor
- Released
- April 2026
- Tokenizer
- Gemma
- Input Modalities
- image, text, video
- Output Modalities
- text
- Max Output
- 32,768 tokens
- Tool Calling
- ✓ supported
- Structured Output
- ✓ supported
- Reasoning Mode
- ✓ supported
- Vision
- ✓ accepts images
- Audio
- no
- Moderated
- no
Category rankings
Where Google: Gemma 4 31B (free) places across the 14 categories it ranks in. How we rank →
| # | Category | Score |
|---|---|---|
| #7 | Cheap Bulk InferenceCost · of 25 ranked | 138 |
| #10 | Self-Hosted / LocalCost · of 25 ranked | 118 |
| #11 | Video SummarizationVideo · of 25 ranked | 150 |
| #15 | Social Media PostsWriting · of 25 ranked | 120 |
| #15 | Voice Assistant BackendVoice · of 25 ranked | 124 |
| #17 | Short-Form SummarizationWriting · of 25 ranked | 129 |
| #17 | Chat CompanionPersonal · of 25 ranked | 129 |
| #19 | Bulk Data LabelingData · of 25 ranked | 133 |
| #19 | Email DraftingWriting · of 25 ranked | 125 |
| #19 | Language LearningEducation · of 25 ranked | 125 |
| #21 | Dataset AnnotationResearch · of 25 ranked | 141 |
| #23 | Real-Time ChatLatency · of 25 ranked | 118 |
| #24 | Customer SupportBusiness · of 25 ranked | 131 |
| #24 | Trivia & General KnowledgePersonal · of 25 ranked | 119 |
Similar models
Google: Gemini 4 Argon
Google: Gemini 4 Argon (batch)
Google: Gemini 3.8 Flash
Google: Gemini 3.8 Flash (batch)
Google: Gemini 3.7 Flash
Google: Gemini 3.7 Flash (batch)
Quick answers
- Is Google: Gemma 4 31B (free) free to use?
- Yes. It is currently listed at no cost per token via OpenRouter, subject to provider rate limits.
- What is Google: Gemma 4 31B (free)'s context window?
- 262,144 tokens, roughly 393 pages of text in a single request.
- Does Google: Gemma 4 31B (free) support tool calling?
- It supports tool calling, structured output, a reasoning mode.
- Can Google: Gemma 4 31B (free) process images?
- Yes, it accepts image input.