Run DeepSeek V4 Flash on M3 Ultra

Yes — DeepSeek V4 Flash (284B-A13B) runs at 30 tok/s on M3 Ultra with 256 GB RAM using Q4_K_M quantization via MLX. First token latency is 0.5s. A capable open-source LLM with 284B-A13B parameters.

Speed
30
tok/s
First Token
0.5
seconds
RAM Needed
256
GB minimum
Engine
MLX
recommended
Want to run DeepSeek V4 Flash faster — or step up to bigger models? Find your Mac →

Benchmark Details

The LLMCheck index estimates DeepSeek V4 Flash on M3 Ultra 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 →

MetricValue
Tokens per second30 tok/s
Time to first token0.5s
QuantizationQ4_K_M
Minimum RAM256 GB
Recommended engineMLX
Parameters284B-A13B
Benchmark date2026-08

Q4_K_M 284B-A13B MLX M3 Ultra

Setup Guide: Run DeepSeek V4 Flash on M3 Ultra

The recommended engine for DeepSeek V4 Flash on M3 Ultra is MLX. Install with pip and pull the model:

pip install mlx-lm
mlx_lm.generate --model mlx-community/deepseek-v4-flash-q4_k_m --prompt "Hello!"

Alternatively, you can use Ollama for a simpler setup:

ollama run deepseek-v4-flash

System Requirements

To run DeepSeek V4 Flash on M3 Ultra you need:

🛒 Get a Mac that runs DeepSeek V4 Flash

DeepSeek V4 Flash needs about 256 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.

Compare More Models

See how DeepSeek V4 Flash stacks up against other models on your specific Mac hardware.

Open Compare Tool Full Leaderboard