Open Source Models
Mixtrain gives you first-class support for open-source models — from robotics foundation models to speech recognition to computer vision. Bring state-of-the-art research into your workflows without vendor lock-in.
Physical AI
Foundation models for robot control and manipulation
Audio & Speech
Speech recognition, text-to-speech, and audio models
Vision
Image understanding, segmentation, and visual reasoning
Why Open Source on Mixtrain?
- No vendor lock-in — run any model from Hugging Face, NVIDIA NGC, or load your own weights.
- Fine-tune on your data — customize models for your specific tasks and hardware.
- Iterate fast — evaluate, compare, and swap models in your workflows without changing a line of code.
- Deploy anywhere — export models to run on your own infrastructure, edge devices, or robots.