· Models

Model families Photon runs.

Every family with production support. Each page lists the architectures that family is validated on.

Open vision-language models for local perception and physical AI.

Open multimodal models in four sizes, compiled through the same path.

Google open multimodal models in two sizes, verified on H100.

OpenAI's open speech-recognition model, served through Photon.

Qwen speech-recognition models in two sizes.

NVIDIA's Parakeet TDT speech-recognition model.