Yes — KAT-Coder-V2.5 (35B-A3B) runs at 47 tok/s on M4 Max with 48 GB RAM using Q4_K_M quantization via MLX. First token latency is 0.3s. A capable open-source LLM with 35B-A3B parameters.
Want to run KAT-Coder-V2.5 faster — or step up to bigger models? Find your Mac →The LLMCheck index estimates KAT-Coder-V2.5 on M4 Max using our published methodology: Q4_K_M quantization, memory-bandwidth scaling, and cross-referenced third-party benchmarks where available. Figures are transparent estimates — own this config? Submit a real benchmark →
| Metric | Value |
|---|---|
| Tokens per second | 47 tok/s |
| Time to first token | 0.3s |
| Quantization | Q4_K_M |
| Minimum RAM | 48 GB |
| Recommended engine | MLX |
| Parameters | 35B-A3B |
| Benchmark date | 2026-08 |
Q4_K_M 35B-A3B MLX M4 Max
The recommended engine for KAT-Coder-V2.5 on M4 Max is MLX. Install with pip and pull the model:
Alternatively, you can use Ollama for a simpler setup:
| Chip | Speed | First Token | Min RAM | Engine |
|---|---|---|---|---|
| M5 Max | 52 tok/s | 0.3s | 64 GB | MLX |
| M4 Pro | 24 tok/s | 0.4s | 24 GB | Ollama |
To run KAT-Coder-V2.5 on M4 Max you need:
KAT-Coder-V2.5 needs about 48 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 our benchmarks free.
See how KAT-Coder-V2.5 stacks up against other models on your specific Mac hardware.
Open Compare Tool Full Leaderboard