gemma-4-26B-A4B-it

Checkpointgoogle/gemma-4-26B-A4B-itModel card indexedModel card

Support summary

gemma-4-26B-A4B-it is a checkpoint published by Google, built on the Gemma4ForConditionalGeneration architecture. vLLM 0.28.0, the newest indexed vLLM release, documents support. Official recipes or cookbooks cover SGLang, without pinning a stable release.

vLLM documents support in 10 of 20 indexed releases, from 0.22.0 to 0.28.0.

Documented hardware
CPU, NVIDIA / CUDA, and AMD / ROCm
Documented quantization
BF16
Documented features
Reasoning parsers and Tool calling
Configured context
262,144 tokens
License
apache-2.0

Hardware, quantization and features count only statements about this checkpoint; engine-wide capabilities are listed with each release.

Hugging Face model card
Publisher
Google
Task
image text to text
License
apache-2.0
Library
transformers
Architecture
Gemma4ForConditionalGeneration
Weight tensor elements
25.8B
Tensor types
BF16
Configured context
262,144 tokens
Base models
google/gemma-4-26B-A4B
Weight access
Public
Repository created
2026-03-11
Source 1: google/gemma-4-26B-A4B-it · publisher model cardSource 2: google/gemma-4-26B-A4B-it · Hugging Face metadataSource 3: google/gemma-4-26B-A4B-it · publisher configurationReviewed
Model notes and sources

Engine support

3 engines

Engine support for gemma-4-26B-A4B-it
EngineReleaseCheckpoint supportEvidence scopeLast known supported
vLLMVersion not pinnedRecipe / cookbookDocumented recipe

Official recipe

Source: vllm · models/Google/gemma-4-26B-A4B-it.yamlReviewed
View 0.28.0
Features, dependencies and evidence for vLLM 0.28.0
SGLangVersion not pinnedRecipe / cookbookDocumented recipe

Official recipe

Source 1: sglang · docs/cookbook/autoregressive/Google/Gemma4.mdxSource 2: sglang · docs/src/snippets/autoregressive/gemma4-deployment.jsxReviewed
No supported release indexed
Features, dependencies and evidence for SGLang 0.5.19
TensorRT-LLM1.2.1Newest indexedNo matching evidence curatedNo supported release indexed
Features, dependencies and evidence for TensorRT-LLM 1.2.1

Recipes and cookbooks