DeepSeek R1 0528
671.2B total, 37.7B active · 160k context · MIT License · released 2025-05-28
chatreasoningcodingagentic
In short
DeepSeek R1 0528 is an open model from DeepSeek with 671.2B parameters (37.7B active per token). At mlx-3bit, DeepSeek R1 0528 needs about 275 GB of memory at 8k context and 277 GB at 32k, counting the weights, the KV cache and runtime buffers. Nothing we price in Canada runs DeepSeek R1 0528 fully in memory at 32k context at a usable speed; see the table for what comes closest.
Updated . Speeds are estimates; how they are worked out.
What to run it on in Canada
Fully in memory at 32k context, priced at retailers today.
Something we price holds it fully in memory, but none reaches 15 tok/s, our bar for recommending it. The table below has the speeds.
All hardware that runs it
Everything we price, best value first. Two cards split the model by layers.
| Hardware | Quant | Fits | Decode | Price | tok/s per 1,000 |
|---|---|---|---|---|---|
| Mac Studio M3 Ultra 512GB | Q4_K_M | fits | 13 tok/s | – | – |
39 other cards and machines we price cannot hold it at this context.
Memory needed
Weights, KV cache and runtime buffers at each context length (KV cache at fp16).
| Quant | File | @8k | @32k | @64k | @128k |
|---|---|---|---|---|---|
| mlx-3bitmlx | 273.6 GB | 274.9 GB | 276.6 GB | 278.9 GB | 283.6 GB |
| Q4_K_Mgguf | 377.1 GB | 378.4 GB | 380.2 GB | 382.5 GB | 387.2 GB |
| mlx-5bitmlx | 429.8 GB | 431.1 GB | 432.8 GB | 435.1 GB | 439.8 GB |
| Q5_K_Mgguf | 443.1 GB | 444.4 GB | 446.1 GB | 448.5 GB | 453.1 GB |
| Q6_Kgguf | 513.2 GB | 514.5 GB | 516.2 GB | 518.6 GB | 523.2 GB |
| Q8_0gguf | 664.3 GB | 665.6 GB | 667.3 GB | 669.7 GB | 674.3 GB |
Run it
Copy a command for the runtime you use.
- mlx-3bit
- LM Studio
lms get mlx-community/DeepSeek-R1-0528-3bit - MLX
mlx_lm.server --model mlx-community/DeepSeek-R1-0528-3bit
- LM Studio
- Q4_K_M
- Ollama
ollama run hf.co/unsloth/DeepSeek-R1-0528-GGUF:Q4_K_M - LM Studio
lms get unsloth/DeepSeek-R1-0528-GGUF@q4_k_m - llama.cpp
llama-server -hf unsloth/DeepSeek-R1-0528-GGUF:Q4_K_M
- Ollama
- mlx-5bit
- LM Studio
lms get mlx-community/DeepSeek-R1-0528-5bit - MLX
mlx_lm.server --model mlx-community/DeepSeek-R1-0528-5bit
- LM Studio
- Q5_K_M
- Ollama
ollama run hf.co/unsloth/DeepSeek-R1-0528-GGUF:Q5_K_M - LM Studio
lms get unsloth/DeepSeek-R1-0528-GGUF@q5_k_m - llama.cpp
llama-server -hf unsloth/DeepSeek-R1-0528-GGUF:Q5_K_M
- Ollama
- Q6_K
- Ollama
ollama run hf.co/unsloth/DeepSeek-R1-0528-GGUF:Q6_K - LM Studio
lms get unsloth/DeepSeek-R1-0528-GGUF@q6_k - llama.cpp
llama-server -hf unsloth/DeepSeek-R1-0528-GGUF:Q6_K
- Ollama
- Q8_0
- Ollama
ollama run hf.co/unsloth/DeepSeek-R1-0528-GGUF:Q8_0 - LM Studio
lms get unsloth/DeepSeek-R1-0528-GGUF@q8_0 - llama.cpp
llama-server -hf unsloth/DeepSeek-R1-0528-GGUF:Q8_0
- Ollama
Notes
Supports system prompts and function calling; no need to force "<think>" at the start of output. Maximum generation length was set to 64K tokens in evaluations. Excludes 13.4B parameters of multi-token-prediction layers, which local runtimes do not load.
Questions
- How much VRAM does DeepSeek R1 0528 need?
- At mlx-3bit, DeepSeek R1 0528 needs about 275 GB of memory at 8k context and 277 GB at 32k, counting the weights, the KV cache and runtime buffers.
- Can I run DeepSeek R1 0528 on a 16 GB GPU?
- No: it does not fit in 16 GB at 32k context, even with system RAM.
- Can I run DeepSeek R1 0528 on a 24 GB GPU?
- No: it does not fit in 24 GB at 32k context, even with system RAM.
- How do I run DeepSeek R1 0528 locally?
- With Ollama: ollama run hf.co/unsloth/DeepSeek-R1-0528-GGUF:Q4_K_M (the Q4_K_M build). LM Studio and llama.cpp commands are on this page.
- What licence is DeepSeek R1 0528 released under?
- MIT License, which allows commercial use.