{"id":"google/gemma-4-26b-a4b-it","aliases":["google/gemma-4-26b-a4b-it-20260403","google/gemma-4-26b-a4b-it:free","hf:google/gemma-4-26b-a4b-it","openrouter:google/gemma-4-26b-a4b-it","openrouter:google/gemma-4-26b-a4b-it:free"],"displayName":"gemma-4-26B-A4B-it","author":"google","createdAt":"2026-03-11T21:25:57.000Z","updatedAt":"2026-09-08T15:44:27.512Z","popularityRank":79,"metadata":{"modalities":["image","image-text-to-text","text","video"],"inputModalities":["image","text","video"],"outputModalities":["text"],"contextLength":262144,"maxOutputTokens":16384,"family":"gemma","modelsDev":{"provider":"openrouter","sourceUrl":"https://models.dev/api.json","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","lastUpdated":"2026-04-02","releaseDate":"2026-04-02","capabilities":{"reasoning":true,"tool_call":true,"attachment":true,"temperature":true,"open_weights":true,"structured_output":true}},"tokenizer":"Gemma","license":"apache-2.0","licenseStatus":"approved","weightsAvailable":true,"hostedInferenceAllowed":true,"gated":false,"supportedParameters":["frequency_penalty","include_reasoning","logit_bias","logprobs","max_tokens","min_p","presence_penalty","reasoning","repetition_penalty","response_format","seed","stop","structured_outputs","temperature","tool_choice","tools","top_k","top_logprobs","top_p"],"providerNames":[],"isModerated":false},"availability":"unknown","externalBest":null,"multivibeNetwork":null,"description":"Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind. Despite 25.2B total parameters, only 3.8B activate per token during inference — delivering near-31B quality at...","runtimes":[],"protocols":[],"harnesses":[]}