The request names gemma-4-26b-a4b-it-mlx-4bit, an MLX 4-bit build of Google's Gemma 4 26B-A4B instruct model meant for local Apple-silicon use. None of the sources is that model's own page: Replicate 1 hosts a Pruna AI-optimised Gemma 4 26B MoE and other Gemma models through a cloud API, Hugging Face 5 hosts open model weights and documents running GGUF models locally, and Ollama 8 runs open models locally on macOS. The request does not state whether the goal is local or cloud use, so the

A question back: Do you want to run this model locally on your Mac, or call it through a cloud API?

1. 1

   Hugging Facehuggingface.coverified and signed

   Hosts open model repositories and documents selecting and running GGUF models locally on CPU, Mac Metal, CUDA or ROCm 5.

   WebsiteDocsPricingGet started

   official:MCP server · @huggingface/mcp-clientSDK · huggingface-hubCLI · text-embeddings-inferencerepository · pwc-cliskill · hf-cli+23 community tools built for it

   for agents: models · model-hub · huggingface-local-models

2. 2

   Ollamaollama.comwebsite read, tools not yet checked

   Runs open AI models locally on macOS, Windows and Linux, and in the cloud 8.

   WebsiteDocsPricingGet started

   +141 community tools built for it

3. 3

   Replicatereplicate.comverified and signed

   Runs open-source models through a cloud API, and lists a Pruna AI-optimised Gemma 4 26B MoE model 1.

   WebsiteDocsPricingGet started

   +14 community tools built for it

   for agents: prunaai/gemma-4-26b-a4b-fast · google-deepmind/gemma-3-4b-it · lucataco/gemma-4-31b-it