Voxtral Mini 3B

This is the edge-deployment member of the Voxtral family — a 3B-parameter speech-understanding model that shares Voxtral Small 24B's architectural approach at a smaller scale. Unlike the transcription-optimized Voxtral Mini Transcribe V2 and the streaming-optimized Voxtral Mini 4B Realtime, this model preserves the full speech-understanding capability (audio Q&A, summarization, audio-grounded reasoning) in a footprint small enough for edge deployment. Use this when you need on-deviceRunning a model directly on a consumer device — a laptop, a phone, a smart speaker — rather than in a data center. On-device inference keeps data private by never leaving the device, and works offline. Small models (under ~10B parameters, often quantized) can run on-device; larger models cannot yet. speech understanding — laptop-class devices, private cloud deployments with limited GPUThe specialized chip that runs most AI models. Originally designed for 3D graphics, GPUs turned out to be excellent at the math AI requires. Nvidia dominates the AI GPU market; common datacenter models include the H100, H200, and B200. Running an AI model without a GPU is possible but painfully slow for anything but the smallest models. budget — rather than pure transcription. As with Voxtral Small 24B, the model can act as a drop-in replacement for its corresponding text-only Mistral base (Ministral 3B in this case) if you need a text-only deployment without maintaining separate checkpoints.