← Back to API Documentation Home

vllm-gemma4-e4b

Gemma 4 E4B BF16 inference via vLLM. Only Gemma 4 model with audio input (30s clips, 16 kHz). Also supports image and video. 128K context.