The LLMCheck index estimates Kimi K2.7 Code at ~27 tok/s on an M5 Ultra at IQ3_S, a figure from its published bandwidth model, not a measured run. It needs about 421 GB of unified memory. It is Moonshot AI's 1T mixture-of-experts model (32B active), released under the Modified MIT licence.
Want to run Kimi K2.7 Code faster — or step up to bigger models? Find your Mac →This is an LLMCheck index estimate from its published two-term bandwidth model, not a measured run; the index labels it so everywhere it appears. Own this Mac? Submit a real run →
| Metric | Value |
|---|---|
| Tokens per second | estimated~27 tok/s |
| Provenance | estimate · how estimates are made |
| Quantization | IQ3_S |
| Engine | llama.cpp |
| Figure dated | 2026-10 |
| Minimum RAM | 421 GB |
| Parameters | 1T MoE (32B active) |
| Chip bandwidth | 1228 GB/s |
Each row as the index holds it, with its source. “Mac RAM” is the memory of the Mac a run used, not the model’s requirement (421 GB). The headline figure follows the reference rule: the lower-median measured plain-decode run at 5 bits or fewer (a run whose quant was not stated counts only when no stated-quant run qualifies), else the estimate; a run marked “not used to anchor estimates” is listed here but never becomes the headline.
| Speed | Quant | Engine | Mac RAM | Context | Source | Note |
|---|---|---|---|---|---|---|
| estimated27 tok/s | IQ3_S | llama.cpp | 512 GB | — | LLMCheck estimator | — |
Ollama — the index has not verified an Ollama tag for Kimi K2.7 Code, so it doesn’t print one (a guessed tag fails with “pull model manifest: file does not exist”). Search the Ollama library →
MLX — pick a build from Hugging Face’s MLX listings, then run it with its repo id:
Downloads: Hugging Face repo → · model card →
Prefer a GUI? LM Studio downloads GGUF and MLX builds in one click. Step by step: install Ollama on a Mac.
Moonshot's coding-focused update to Kimi K2.6 (Jun 2026, Modified MIT): a 1T MoE with 32B active, 256K context and image input; no SWE-bench figure is published. It fits only a 512 GB Mac Studio, at ~3–3.6-bit (419–460 GB); the native INT4 release (~595 GB) fits no Mac.
Modified MIT licence · 1T MoE · 256K context · LLMCheck capability 48/50 · see its leaderboard rank →
Kimi K2.7 Code needs about 421 GB of unified memory. These current Apple Silicon Macs have the headroom to run it comfortably:
As an Amazon Associate, LLMCheck earns from qualifying purchases. These affiliate links cost you nothing extra and help keep the LLMCheck index free.
See how Kimi K2.7 Code stacks up against other models on your specific Mac hardware.
Open Compare Tool Full Leaderboard