floneum/kalosm

Health v2 · maintenance only

Adoption not part of the score

2223 stars · 133 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Kalosm is a Rust ecosystem of crates providing simple interfaces for running pre-trained language, audio, and image models locally or remotely. It includes Fusor, a WGPU-based runtime for quantized ML inference on accelerators or in the browser.

Use cases

• run local llm chat in rust
• transcribe audio with whisper locally
• build a private rag chatbot over documents
• semantic search with bert embeddings
• segment images with segment anything
• run quantized models in the browser with wgpu
• constrained structured text generation

When to choose

• you want private, local AI inference in a Rust application
• you need multi-modal support (text, audio, image) from one library
• you want GPU-accelerated quantized inference without Python
• you need retrieval-augmented generation with built-in vector storage

When to avoid

• you need the absolute latest models or broadest model ecosystem (Python tooling is richer)
• your stack is not Rust
• you need managed cloud inference rather than local models

Facets

library · maturity active

machine-learning llm-inference speech-recognition rag chatbot sdk artificial-intelligence large-language-models machine-learning speech-processing rust cross-platform wasm local-ai llamacpp candle whisper constrained-generation wgpu embeddings semantic-search image-segmentation natural-language-processing retrieval-augmented-generation gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="floneum/kalosm")