The LLMCheck index estimates Recursive Flint at ~131 tok/s on an M5 Max at Q4_K_M, scaled from its community-measured M3 Ultra run with the published bandwidth model, not a measured run on this chip. It needs about 20 GB of unified memory. It is Recursive's 39.8B mixture-of-experts model (2.5B active), released under the Apache 2.0 licence.
Want to run Recursive Flint 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~131 tok/s |
| Provenance | estimate · how estimates are made |
| Quantization | Q4_K_M |
| Engine | llama.cpp |
| Figure dated | 2026-10 |
| Minimum RAM | 20 GB |
| Parameters | 39.8B MoE (2.5B active) |
| Chip bandwidth | 614 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 (20 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 |
|---|---|---|---|---|---|---|
| community131 tok/s | INT4 | MLX | 36 GB | ~1K context, greedy, 256 generated | github.com | The 36 GB M5 Max is the binned 32-core-GPU, 460 GB/s part, so this run never anchors or sets the reference. Rapid-MLX 0.16.8 on MLX 0.32.4, RecursiveSI/flint-int4; speculative decoding and MTP off per the release notes; ~24 GB peak. 118.5 tok/s at ~8K. Published by the runtime's developers. 4.54 bpw = the 22.6 GB RecursiveSI/flint-int4 checkpoint over 39.8B parameters.; binned GPU |
| community110.2 tok/s | oQ6e | MLX | 128 GB | 1k | omlx.ai | 40-core GPU. oMLX v0.7.2+omni (an unexplained build), MTP/DFlash off; Flint-oQ6e, a format the index cannot cost, so it does not anchor; 30.8 GB peak. 102.5 tok/s at 4k.; not used to anchor estimates |
| estimated131 tok/s | Q4_K_M | llama.cpp | 64 GB | — | LLMCheck estimator | — |
Ollama — install it (brew install ollama or the app from ollama.com), then pull the model by its tag, checked on ollama.com on 2026-10-24:
MLX — pick a build from Hugging Face’s MLX listings, then run it with its repo id:
Downloads: Ollama page → · model card →
Prefer a GUI? LM Studio downloads GGUF and MLX builds in one click. Step by step: install Ollama on a Mac.
Recursive's Apache 2.0 coding-agent MoE (21 Oct 2026): 39.8B total, ~2.5B active, 262K context, text only, always reasons. Its card reports SWE-bench Verified 74.2 and Terminal-Bench 2.1 54.9 (its SWE-bench Pro v2 89.7 is an internal mini-swe-agent run, not comparable to SEAL's); the index's capability figure is estimated from these vendor coding benchmarks only. Stock llama.cpp b11803+, Ollama (flint:40b) and LM Studio run it, though llama.cpp's expert-scaling workaround costs 10–30% for now and issue #31000 (NVFP4 tool calls) is open. A ~18 GB 3-bit build fits a 24 GB Mac; community llama.cpp runs measured 92.4 tok/s on an M5 Pro and 79.6 on an M4 Pro (Q4_1).
Apache 2.0 licence · 39.8B MoE · 262K context · LLMCheck capability 37/50 · see its leaderboard rank →
Recursive Flint needs about 20 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 Recursive Flint stacks up against other models on your specific Mac hardware.
Open Compare Tool Full Leaderboard