Google DeepMind
Google DeepMind
Language model · United States ·Jun 2026

DiffusionGemma 26B-A4B

Language Open Weights
26B
Parameters

About DiffusionGemma 26B-A4B

DiffusionGemma 26B-A4B is an AI model developed by Google DeepMind, in the language category, released in 2026, made available as an open-weights model with 26B params. DiffusionGemma is Google DeepMind's experimental 26B-parameter (3.8B active) Mixture-of-Experts open model that uses tex.

On this page you'll find DiffusionGemma 26B-A4B's full specifications. Review provider pricing and benchmark scores below, or compare DiffusionGemma 26B-A4B head-to-head with other language models.

→ More models by Google DeepMind → All language models → AI benchmarks leaderboard

Run DiffusionGemma 26B-A4B locally — what to buy

Share this build ↗

Sized to what DiffusionGemma 26B-A4B actually needs (~8k context at Q4), not to the biggest GPU: Good = cheapest that runs it, Better = best value, Best = most headroom. Speed is a hardware estimate; anything past ~120 tok/s is shown as instant.

Good
AMD Radeon RX 7900 XT
Consumer GPU · ~$899
~42.3 tok/s
Q4 · fits full context · up to 8k ctx
Runs with: ExLlamaV2, KoboldCpp, LM Studio
See full estimate →
Better
AMD Radeon RX 7900 XTX
Consumer GPU · ~$999
~50.8 tok/s
Q4 · fits full context · up to 8k ctx
Runs with: ExLlamaV2, KoboldCpp, LM Studio
See full estimate →
Best
NVIDIA GeForce RTX 5090
Consumer GPU · ~$1,999
~107.1 tok/s
Q4 · fits full context · up to 8k ctx
Runs with: ExLlamaV2, KoboldCpp, LM Studio
See full estimate →

Software support

·ExLlamaV2 ·KoboldCpp ·LM Studio ·MLX ·Ollama ·SGLang ·TensorRT-LLM ·llama.cpp ·vLLM ?KTransformers ?text-generation-webui

✓ measured · · compatible · — not supported. Informational only — speed is hardware-based.

Consumer GPU
13.5–107.1 tok/s
10 options →
Apple
16.9–45.1 tok/s
17 options →
Datacenter GPU
45.7–120+ tok/s
13 options →

Community

Would you run DiffusionGemma 26B-A4B again?
No votes yet — be the first.
Did it run? — community reports by hardware

No reports yet. Own this model on some hardware? Be the first to confirm it runs.

Discussion

No comments yet. Start the discussion.

More models like DiffusionGemma 26B-A4B

AI
DMind-1
 AI
COLLECTIVE-1
 AI
Cogito 1 series
 AI
Euler (Euler大模型)
· 2.5B
 AI
SmolLM-1.7B
· 1.7B
 AI
Llama-3-70B-Synthia-v3.5
· 70B

Frequently asked questions

What is DiffusionGemma 26B-A4B? ▶

DiffusionGemma is Google DeepMind's experimental 26B-parameter (3.8B active) Mixture-of-Experts open model that uses tex It is tracked on GenAIList with its specifications, benchmark scores and provider pricing.

Who created DiffusionGemma 26B-A4B? ▶

DiffusionGemma 26B-A4B was developed by Google DeepMind and released in 2026.

Is DiffusionGemma 26B-A4B open source or proprietary? ▶

DiffusionGemma 26B-A4B ships as an open-weights model: you can download and self-host the weights, though the licence may place some restrictions on use.

How much does DiffusionGemma 26B-A4B cost? ▶

Pricing for DiffusionGemma 26B-A4B depends on the provider. See the providers table on this page for the latest API rates.

How does DiffusionGemma 26B-A4B perform on benchmarks? ▶

Benchmark scores for DiffusionGemma 26B-A4B are listed on this page as they are published. You can compare it head-to-head with other models on the GenAIList compare tool.

DiffusionGemma 26B Review: 4x Faster, Real Tradeoffs
DiffusionGemma Local Test | 4x Faster but How Accurate?
Diffusion Gemma: Google's First Open Diffusion Model
Google's New AI Doesn't Type. It Develops | DiffusionGemma
Diffusion Gemma First Look & Demo – A BIG Step for Local AI Models!

Know a great video about DiffusionGemma 26B-A4B?