Kapitsa is an inference engine for Apple silicon that runs Qwen and Nemotron models locally, built for coding agents. It starts opencode, Claude Code, Codex or pi on a local model, and serves OpenAI and Anthropic APIs.

Quickstart · CLI · Benchmarks · Models · How KFlash works · Docs

• [2026/10] 🚀 KFlash: tree speculative decoding, up to 3.5× llama.cpp and nearly 2× oMLX. One verify pass checks a whole tree of the drafter's guesses. M4 Pro, decode at T=0.6: 63 vs 18 tok/s on the 27B, 201 vs 102 on the 4B; +6% over the build before it on the 27B, −3% on a long agent turn (how it works, #187).
• [2026/10] 🏆 Kapitsa: the fastest decode in 21 of 24 model and Mac pairs against eight engines, up to 2.3× MLX (4B, M4 Pro, 201 vs 88 tok/s). On the M5 Pro Kapitsa is the 10-21 re-run and the others the 10-19 runs (poster).
• [2026/10] 🤝 KFlash budget: subagents decode together. Three requests at once share one verify pass, up to +23% total throughput on the 27B (#207).
• [2026/10] ⚡ QMT prefill: read the prompt up to 1.5× faster on M5. Metal 4 matrix units for the prompt (Qwen3.5 4B, M5 Pro, a busy GPU, not bit-identical, no output check yet; #205, #215).
• [2026/10] 💾 SSD prefix cache: restart the server, keep the conversation. M4 Pro, 35B: a 16K-token follow-up starts in 0.16 s instead of 17 s, about 100× (#212, --disk-cache).
• [2026/10] 🧠 Prefix reuse: subagents start warm. A new question over the shared system prompt answers in 0.06 s instead of 1.9 s (4K, 4B, M5 Pro, a busy GPU; #198, #202).
• [2026/10] 🧪 Output check: speed has to prove it kept the answers. From now on a default changes only after its logits stay within twice the model's noise floor (how, #208, #210).
• [2026/10] 📊 kapitsa bench compare: see it on your own Mac. The same test against every engine installed here, in one command (#197, #219).

Requirements: a Mac with Apple silicon (M1 or newer) and macOS 14 Sonoma or later.

brew install nace-ai/tap/kapitsa
# or: curl -fsSL https://raw.githubusercontent.com/nace-ai/kapitsa/main/install.sh | bash

kapitsa opencode --model 35b    # also: claude, codex, pi

_{On a 16-24 GB Mac use --model 9b, or 4b for the quickest replies.}

kapitsa serve                   # chat + OpenAI/Anthropic APIs at http://127.0.0.1:8000
kapitsa claude                  # Claude Code on Kapitsa
kapitsa doctor                  # this Mac's GPU and memory, and which models fit it
kapitsa bench                   # benchmark this Mac, share it with the community
kapitsa bench compare           # the same test against the other engines installed here
kapitsa pull 27b                # download a model without starting it
kapitsa list                    # downloaded models and their size
kapitsa rm 27b                  # delete a downloaded model
kapitsa import model.gguf       # use a Qwen GGUF from LM Studio, Ollama or llama.cpp
kapitsa update                  # update now (it also updates itself)

_{--model picks the model and its precision (35b:6bit). kapitsa <command> --help has the options. Usage guide}

Kapitsa has the fastest decode in 21 of the 24 model and Mac pairs measured against eight engines (on the M5 Pro, Kapitsa was re-run on 10-21 and the other engines are the 10-19 runs, except mlx-serve's 27B).

_{Each cell reads prefill / decode in tokens per second. Prefill is a 4K-token prompt, decode a 512-token answer. × cannot run that model · — not measured. How these were measured, run dates and flags: docs/benchmarks.md.}

Every recorded run, with its Mac, model, prefill and decode speed, is in the benchmark table. Measure your Mac: kapitsa bench --submit adds it to the community results. Another engine faster on your Mac? kapitsa bench compare ranks every engine you have installed, and --report writes a Markdown report.

Kapitsa decodes with KFlash. A draft model such as DFlash2 guesses the next seven tokens, with a few candidates for each one. Engines that run it as published (Splash, oMLX, mlx-serve) check one chain of guesses and stop at the first wrong one. KFlash checks a best-first tree of the drafter's candidates in the same pass, so where the drafter hesitates, its runner-ups are checked too and more tokens survive each pass. The tree's size comes from what a verify row costs on your Mac, measured there, and it changes no output: the big model checks every token. It runs on the published drafters as they are: z-lab's DFlash, DFlash2, and NVIDIA's DSpark and LiLiCorr. How KFlash works

_{Watch → Custom → Releases on nace-ai/kapitsa puts every new version in your notifications. A drawn illustration of the steps, not a screen recording; the page source is in tools/readme_images.}

_{How KFlash works · Architecture · Contributing · Press kit · Apache 2.0, © 2026 Nace AI · Third-party notices}