Intelligence Index
0129.7
> 46% OF MODELS
Google
Gemma 4 31B
Updated: September 2026
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...
Specifications
- Context
- 262K
- Input Modalities
- Text, Image
- Output Modalities
- Text
- Input
- $0.09/M
- Output
- $0.34/M
Capabilities
VISIONTEXTWEBCODINGTHINKINGWRITING
Similarly Priced Models
| Model | Provider | Context | Input Price | Output Price |
|---|---|---|---|---|
| Qwen3 235B A22B Instruct 2507 | Qwen | 262K | $0.0875/M | $0.35/M |
| Qwen3 32B | Qwen | 131K | $0.08/M | $0.28/M |
| Gemma 3 27B | Google | 131K | $0.08/M | $0.45/M |
| GPT-5.6 Luna | OpenAI | 1M | $0.1/M | $0.6/M |
| GPT-5.6 Luna | OpenAI | 1M | $0.1/M | $0.6/M |
Performance Metrics
Coding Index
0243.4
> 43% OF MODELS
Agentic Index
0314.4
> 32% OF MODELS
Average Response Performance
- Output Speed
- 35.1 tok/s
- Time To First Token
- 1.11s
- Time To First Answer Token
- 50.53s
- End To End Response Time
- 64.76s
DATA SOURCE: Artificial Analysis