Gemma 4 26B A4B
AffordableRun locallyEfficient Mixture-of-Experts Gemma 4 with only 4B active parameters per token.
26B total parameters delivering near-31B quality at a fraction of the compute. Supports text, image, and video inputs with 262K context and configurable reasoning.
▼
Gemma 4 26B A4B is somewhat expensive at $0.07/M tokens input and $0.34/M tokens output. The model supports vision/image input, extended reasoning, a 262K context window.
Benchmarks
This model hasn't been evaluated by Artificial Analysis yet. Benchmark scores will appear once they're published.
Specifications
Performance
Providers
Run locally
Run Gemma 4 26B A4B on your own computer with the Local AI Engine for free, with no rate limits and no third-party AI provider involved. idapt installs it and routes chat to your hardware automatically.
ollama pull gemma4:26b
Misc
Compare with
GPT 5.6 Solvs Gemma 4 26B A4BGPT 5.6 Lunavs Gemma 4 26B A4B
Claude Fable 5vs Gemma 4 26B A4BClaude Opus 4.8vs Gemma 4 26B A4BClaude Sonnet 5vs Gemma 4 26B A4BClaude Haiku 4.5vs Gemma 4 26B A4B
Gemini 3.6 Flashvs Gemma 4 26B A4BGemini 3.5 Flashvs Gemma 4 26B A4BGemini 3.5 Flash-Litevs Gemma 4 26B A4BGemma 4 31Bvs Gemma 4 26B A4B
Grok 4.5vs Gemma 4 26B A4BGrok Build 0.1vs Gemma 4 26B A4BGrok 4.1 Fastvs Gemma 4 26B A4B
Muse Spark 1.1vs Gemma 4 26B A4B
DeepSeek V4 Provs Gemma 4 26B A4BDeepSeek V4 Flashvs Gemma 4 26B A4B
MiniMax M3vs Gemma 4 26B A4B
GLM 5.2vs Gemma 4 26B A4B
Qwen3.7 Maxvs Gemma 4 26B A4BQwen3.7 Plusvs Gemma 4 26B A4BQwen3 Coder Nextvs Gemma 4 26B A4B
Nemotron 3 Ultravs Gemma 4 26B A4B
Mistral Small 4vs Gemma 4 26B A4BFrequently Asked Questions about Gemma 4 26B A4B▼
When was Gemma 4 26B A4B released?
▼
Gemma 4 26B A4B was released on April 3, 2026.
Who created Gemma 4 26B A4B?
▼
Gemma 4 26B A4B was created by Google.
How much does Gemma 4 26B A4B cost?
▼
Gemma 4 26B A4B costs $0.07/M input tokens and $0.34/M output tokens.
What is Gemma 4 26B A4B API pricing?
▼
The Gemma 4 26B A4B API is priced at $0.07 per million input tokens and $0.34 per million output tokens. You can access Gemma 4 26B A4B through idapt.app alongside 200+ other AI models in one workspace.
Is Gemma 4 26B A4B a reasoning model?
▼
Yes, Gemma 4 26B A4B is a reasoning model. It can think step-by-step through complex problems before providing an answer, often yielding better results on difficult tasks like math, coding, and multi-step logic.
Does Gemma 4 26B A4B support image or vision input?
▼
Yes, Gemma 4 26B A4B supports vision/image input. You can share images in your conversation and the model will analyze and respond based on their content.
Does Gemma 4 26B A4B support audio?
▼
No, Gemma 4 26B A4B does not support audio input.
What is the context window of Gemma 4 26B A4B?
▼
Gemma 4 26B A4B has a context window of 262K tokens, meaning it can process approximately 196,608 words of context at once.
Is Gemma 4 26B A4B open source?
▼
Gemma 4 26B A4B is a proprietary model developed by Google. The model weights and training data are not publicly available.
What is Gemma 4 26B A4B good at?
▼
Based on benchmark data, Gemma 4 26B A4B excels at: visual analysis and image understanding. It is Google's capable model suitable for a wide range of tasks.
Where can I use Gemma 4 26B A4B?
▼
You can chat with Gemma 4 26B A4B on idapt.app — no separate API key required. idapt is an AI workspace with 200+ models, agents, computers, and tasks. You can also access Gemma 4 26B A4B through Google's own API or platform.