Free Gemma model via API

Open Gemma models for experiments and light production. This page is separate from Gemini: it is a different model line.

19 models5 providersup to 262K contextbest: Gemma 4 26B A4BUpdated: October 24, 2026 at 10:40 AM UTC

Gemma does not inherit Google AI Studio Gemini quotas. Read the limit on the selected host.

#	Model ↕	AA Index ↕	GPQA ↕	Context ↕	Size ↕	Free at ↕
1	Gemma 4 26B A4B Google DeepMind · 26B · 4B VisionTool useOpen weights	16.7	79%	262K	Mid-size	OpenRouter, Together AI, Cloudflare Workers AI, UnoRouter
2	Gemma 4 31B Google DeepMind · 31B VisionTool useOpen weights	14.7	86%	262K	Mid-size	OpenRouter, Together AI, NVIDIA Build, UnoRouter
3	Gemma 4 12B Google DeepMind · 12B VisionTool useOpen weights	14.2	75%	262K	Mid-size	Together AI
4	DiffusionGemma 26B Google DeepMind · 26B · 4B Tool useOpen weights	9.5	67%	262K	Mid-size	UnoRouter
5	Gemma 3 27B Google DeepMind · 27B VisionTool useOpen weights	—	—	66K	Mid-size	Together AI
6	Medgemma 27B Text Google DeepMind · 27B Tool useOpen weights	—	—	131K	Mid-size	Together AI
7	Gemma SEA-LION V4 27B Google DeepMind · 27B Tool useOpen weights	—	—	128K	Mid-size	Cloudflare Workers AI
8	Gemma 3 12B Google DeepMind · 12B VisionTool useOpen weights	—	—	—	Mid-size	NVIDIA Build
9	Gemma 4 E4b Google DeepMind VisionOpen weights	8.9	58%	131K	Small	Together AI
10	Gemma 4 E2b Google DeepMind VisionOpen weights	7.8	43%	131K	Small	Together AI
11	Gemma 3 4B Google DeepMind · 4B VisionOpen weights	—	—	66K	Small	Together AI, NVIDIA Build
12	Gemma 3 270M Google DeepMind · 270M Open weights	5.1	22%	33K	Small	Together AI
13	Gemma 3 1B Google DeepMind · 1B Open weights	—	—	33K	Small	Together AI
14	Gemma 7B Google DeepMind · 7B Open weights	—	—	4K	Small	Cloudflare Workers AI
15	Gemma 2B Google DeepMind · 2B · Legacy Open weights	—	—	8K	Small	Together AI, Cloudflare Workers AI, NVIDIA Build
16	Gemma 2 9B Google DeepMind · 9B · Legacy Open weights	—	—	8K	Small	Together AI
17	CodeGemma 1.1 7B Google DeepMind · 7B · Legacy CodingOpen weights	—	—	—	Small	NVIDIA Build
18	CodeGemma 7B Google DeepMind · 7B · Legacy CodingOpen weights	—	—	—	Small	NVIDIA Build
19	RecurrentGemma 2B Google DeepMind · 2B · Legacy Open weights	—	—	—	Small	NVIDIA Build

Benchmarks: Artificial Analysis (October 24, 2026). A dash means the model is not on their leaderboard; we never estimate scores ourselves.

About Gemma

Google DeepMind · open weights

Gemma is Google DeepMind's family of lightweight open-weight models derived from Gemini research, including code and multimodal variants.

Official site

FAQ

Which Gemma models are free?

Currently 19 models are free: Gemma 4 26B A4B, Gemma 4 31B, Gemma 4 12B, DiffusionGemma 26B, Gemma 3 27B, Medgemma 27B Text.

Where can I get a free Gemma API key?

Free Gemma routes are available at: OpenRouter, Together AI, Cloudflare Workers AI, UnoRouter, NVIDIA Build. Get a key in the provider's dashboard; links are on each model card.

Who develops Gemma?

Gemma is Google DeepMind's family of lightweight open-weight models derived from Gemini research, including code and multimodal variants.

About Gemma

FAQ

See also