FFAI
FFAI is the go-to media library for AI: embedded speech recognition (ASR), text-to-speech (TTS), optical character recognition (OCR), vision-language (VLM), and object detection (YOLO26) made in Rust for the new internet.
- 6
- Rust
- 15 h ago
FFAI is the go-to media library for AI: embedded speech recognition (ASR), text-to-speech (TTS), optical character recognition (OCR), vision-language (VLM), and object detection (YOLO26) made in Rust for the new internet.
Speech recognition and text-to-speech in pure Rust: Whisper + WhisperX-class ASR and VITS/Piper-class TTS, no Python, no C/C++ by default, nothing GPL. Memory safe. The ffai-mercury crate.
Object detection in pure Rust: YOLO26 on candle from real Ultralytics. Identical to PyTorch at 3.71x less CPU and a third the memory. ByteTrack tracking and a WebAssembly build. No Python, ONNX, or vendored weights. The ffai-diana crate.
OCR in pure Rust: documents, screens, and a change-gated LIVE frame reader. Detection, recognition and reading order computed from geometry, no Python, no C/C++ by default, nothing GPL. Memory safe. The ffai-carmenta crate.
Vision-language captioning and VQA in pure Rust: SmolVLM-256M on candle — SigLIP tower, pixel-shuffle connector, Llama decoder. Captions byte-identical to the Python reference from a raw image file. Images and video, no Python, no ONNX, nothing GPL. Memory safe. The ffai-argus crate.