Gemma 4 Models by Google DeepMind
Google DeepMindApache 2.0
8 models2026Up to 256K ctxFrom $0.06/1M input
About
Google's most capable open-source model family, purpose-built for advanced reasoning and agentic workflows. Delivered in four sizes (E2B, E4B, 26B MoE, 31B dense) with multimodal capabilities including text, image, video, and audio processing for the smaller models.
Current Variants
Use-when guidance is derived from seed capabilities, context, release, and replacement fields.
8 in view
Gemma 4 E2BCurrent
Use when the workload needs 128K context, 2B parameters, and function calling.
2026-03128K context2B parametersfunction calling
Gemma 4 E2B ITCurrent
Use when the workload needs 128K context, 2B parameters, and function calling.
2026-03128K context2B parametersfunction calling
Gemma 4 E4BCurrent
Use when the workload needs 128K context, 4B parameters, and function calling.
2026-03128K context4B parametersfunction calling
Gemma 4 E4B ITCurrent
Use when the workload needs 128K context, 4B parameters, and function calling.
2026-03128K context4B parametersfunction calling
Gemma 4 26B A4BCurrent
Use when the workload needs 256K context, 26B parameters, and function calling.
2026-03256K context26B parametersfunction calling
Gemma 4 26B A4B ITCurrent
Use when the workload needs 256K context, 26B parameters, and function calling.
2026-03256K context26B parametersfunction calling
Gemma 4 31BCurrent
Use when the workload needs 256K context, 31B parameters, and function calling.
2026-03256K context31B parametersfunction calling
Gemma 4 31B ITCurrent
Use when the workload needs 256K context, 31B parameters, and function calling.
2026-03256K context31B parametersfunction calling
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| Gemma 4 E2B | Use when the workload needs 128K context, 2B parameters, and function calling. | 2026-03 | 128K context2B parametersfunction calling | Current |
| Gemma 4 E2B IT | Use when the workload needs 128K context, 2B parameters, and function calling. | 2026-03 | 128K context2B parametersfunction calling | Current |
| Gemma 4 E4B | Use when the workload needs 128K context, 4B parameters, and function calling. | 2026-03 | 128K context4B parametersfunction calling | Current |
| Gemma 4 E4B IT | Use when the workload needs 128K context, 4B parameters, and function calling. | 2026-03 | 128K context4B parametersfunction calling | Current |
| Gemma 4 26B A4B | Use when the workload needs 256K context, 26B parameters, and function calling. | 2026-03 | 256K context26B parametersfunction calling | Current |
| Gemma 4 26B A4B IT | Use when the workload needs 256K context, 26B parameters, and function calling. | 2026-03 | 256K context26B parametersfunction calling | Current |
| Gemma 4 31B | Use when the workload needs 256K context, 31B parameters, and function calling. | 2026-03 | 256K context31B parametersfunction calling | Current |
| Gemma 4 31B IT | Use when the workload needs 256K context, 31B parameters, and function calling. | 2026-03 | 256K context31B parametersfunction calling | Current |
Release Timeline
1 release group2026-03
8 current
Gemma 4 26B A4B
Current256K context26B parametersfunction calling
Gemma 4 26B A4B IT
Current256K context26B parametersfunction calling
Gemma 4 31B
Current256K context31B parametersfunction calling
Gemma 4 31B IT
Current256K context31B parametersfunction calling
Gemma 4 E2B
Current128K context2B parametersfunction calling
Gemma 4 E2B IT
Current128K context2B parametersfunction calling
Gemma 4 E4B
Current128K context4B parametersfunction calling
Gemma 4 E4B IT
Current128K context4B parametersfunction calling
Specifications(8 models)
| Model | Released | Context | Parameters | Multimodal | Fn Calling | Structured Outputs |
|---|---|---|---|---|---|---|
| Gemma 4 E2B | 2026-03 | 128k | 2B | Yes | Yes | No |
| Gemma 4 E2B IT | 2026-03 | 128k | 2B | Yes | Yes | Yes |
| Gemma 4 E4B | 2026-03 | 128k | 4B | Yes | Yes | No |
| Gemma 4 E4B IT | 2026-03 | 128k | 4B | Yes | Yes | Yes |
| Gemma 4 26B A4B | 2026-03 | 256k | 26B | Yes | Yes | No |
| Gemma 4 26B A4B IT | 2026-03 | 256k | 26B | Yes | Yes | Yes |
| Gemma 4 31B | 2026-03 | 256k | 31B | Yes | Yes | No |
| Gemma 4 31B IT | 2026-03 | 256k | 31B | Yes | Yes | Yes |
Available From(4 providers)
Pricing
| Model | Provider | Input / 1M | Output / 1M | Type |
|---|---|---|---|---|
| Gemma 4 26B A4B IT | OpenRouter | $0.06 | $0.33 | Serverless |
| Gemma 4 31B IT | OpenRouter | $0.13 | $0.38 | Serverless |
| Gemma 4 26B A4B IT | GCP Vertex AI | $0.15 | $0.6 | Serverless |
| Gemma 4 31B IT | GCP Vertex AI | $0.15 | $0.6 | Serverless |
| Gemma 4 31B IT | Together AI | $0.2 | $0.5 | Serverless |
Frequently Asked Questions
- What is Gemma 4 used for?
- Gemma 4 is used for vision and multimodal work, agent workflows and tool use, and structured outputs. The family description and listed model capabilities point to those workloads as the best fit.
- How does Gemma 4 compare to Gemini 3?
- Gemma 4 by Google DeepMind is strongest where you need vision and multimodal work, while Gemini 3 by Google DeepMind is the closest related family to check for robotics. Gemma 4 has 8 listed variants and reaches up to 256K context, while Gemini 3 reaches up to 1M context, so compare the specs and pricing tables before choosing a production model.
- Which Gemma 4 model should I use?
- For the lowest listed input price, start with Gemma 4 26B A4B IT through OpenRouter at $0.06/1M input tokens. For the most capable/latest local choice, evaluate Gemma 4 26B A4B IT with 256K context and function calling, structured outputs, and multimodal inputs.
Related Model Families
Models(8)
Gemma 4 E2B
2026-03128k2B1 provider
MultimodalOpen Source
Gemma 4 E2B IT
2026-03128k2B2 providers
MultimodalOpen Source
Gemma 4 E4B
2026-03128k4B1 provider
MultimodalOpen Source
Gemma 4 E4B IT
2026-03128k4B2 providers
MultimodalOpen Source
Gemma 4 26B A4B
2026-03256k26B
MultimodalOpen Source
Gemma 4 26B A4B IT
2026-03256k26B4 providers
MultimodalOpen Source
Gemma 4 31B
2026-03256k31B
MultimodalOpen Source
Gemma 4 31B IT
2026-03256k31B4 providers
MultimodalOpen Source