kapitsa alternatives

nace-ai/kapitsa (698 stars, +208 this week): Blazing fast Apple Silicon inference engine optimized for Qwen & Nemotron. OpenAI + Anthropic API compatible. Built by folks from nace.ai.

Below are 6 other repos tracked by OSS Radar that share GitHub topics with kapitsa: 5 from LLM tooling sharing at least two about what they do, and 1 from other categories sharing at least three topics about what they do (platform, language and framework tags don't count), two of them niche rather than broad ones like llm, ordered by how closely their topics overlap (rarer shared topics count more), then by stars gained in the last 7 days. Topic overlap means a similar problem space, not a drop-in replacement, so check each project's README.

1. 1. omlxjundot/omlx

   +184/wk

   LLM inference server with continuous batching & SSD caching for Apple Silicon — managed from the macOS menu bar.

   23k stars · Python · shared topics: apple silicon, macOS, MLX, OpenAI API · omlx alternatives

2. 2. HyperQwensyv-ai/HyperQwen

   +138/wk

   Serve large Qwen models fast on the GPUs you actually own. Qwen3.8-27B on a single 24 GB card with vLLM…

   2.6k stars · Python · shared topics: LLM inference, local LLM, Qwen, speculative decoding · HyperQwen alternatives

3. 3. PolyStratasimplexai-labs/PolyStrata

   +155/wkest.

   Run large AI models on your own PC: Qwen3.8-Flash-Next on Strata's engine, and GLM 5.3 Flash, DeepSeek V4 Flash, LFM2.5, Qwen3.6, Ornith…

   918 stars · C++ · shared topics: local LLM, OpenAI API, Qwen, speculative decoding · PolyStrata alternatives

4. 4. RustingStrataRustingStudio/RustingStrata

   +139/wkest.

   Pure-Rust MoE inference engine for Qwen3.5/3.6/3.8 GGUF. CPU+GPU hybrid experts, NVRTC CUDA kernels, OpenAI-compatible server.

   139 stars · Rust · Local LLMs · shared topics: llama.cpp, local LLM, OpenAI API, Qwen · RustingStrata alternatives

5. 5. project-mayamw00/project-maya

   +292/wk

   Run GLM-5.3-Flash (321B), DeepSeek-V4.1-Flash (748B) and MiniMax-M3.1-Flash (428B) on your own GPUs - one engine, one dashboard…

   2.7k stars · Cuda · shared topics: LLM inference, local LLM, OpenAI API · project-maya alternatives

6. 6. SoupMakazhanAlpamys/Soup

   +407/wk

   Fine-tune LLMs from one YAML. Layer streaming trains an 8B model on a 4 GB laptop GPU.

   10k stars · Python · shared topics: local LLM, Ollama · Soup alternatives

More: all trending LLM tooling repos · trending coding agents repos · trending macOS repos · trending Ollama repos · trending Qwen repos · compare kapitsa with the top 2 · alternatives to other trending repos