Yes — Nemotron Cascade 2 (30B) runs at 35 tok/s on M5 Max with 64 GB RAM using Q4_K_M quantization via Ollama. First token latency is 0.9s. NVIDIA's Nemotron Cascade 2, a hybrid SSM-Transformer 30B model.
LLMCheck measured Nemotron Cascade 2 on M5 Max using the standard methodology: Q4_K_M quantization, 256-token input, 512-token output, 3 runs averaged on a freshly-booted system.
| Metric | Value |
|---|---|
| Tokens per second | 35 tok/s |
| Time to first token | 0.9s |
| Quantization | Q4_K_M |
| Minimum RAM | 64 GB |
| Recommended engine | Ollama |
| Parameters | 30B |
| Benchmark date | 2026-04 |
Q4_K_M 30B Ollama M5 Max
The recommended engine for Nemotron Cascade 2 on M5 Max is Ollama. Install Ollama, then pull the model:
Ollama handles quantization automatically — it will download the Q4_K_M variant (~64 GB) and start an interactive chat session.
| Chip | Speed | First Token | Min RAM | Engine |
|---|---|---|---|---|
| M4 Max | 28 tok/s | 1.2s | 48 GB | MLX |
| M4 Pro | 22 tok/s | 1.5s | 24 GB | Ollama |
To run Nemotron Cascade 2 on M5 Max you need:
See how Nemotron Cascade 2 stacks up against other models on your specific Mac hardware.
Open Compare Tool Full Leaderboard