DeepSeek Coder

What DeepSeek Coder is for

DeepSeek Coder is trained for code. DeepSeek released it in 2023-11 with 3 sizes (1.3B, 6.7B, 33B), 16K of context and a DeepSeek License licence. Context length matters more here than for chat, because a coding assistant that cannot see the whole file is guessing. The catalogue records its strength as small-footprint code completion, and it is the class of model an editor plugin such as Continue or Tabby points at.

Sizes and memory

The 1.3B needs about 1 GB of weights at Q4_K_M and the 33B about 20 GB, a spread of 19 GB across the family. The catalogue's default pick is 6.7B. Recommended VRAM adds roughly half again on top of the weights for the KV cache and the runtime.

Size	Weights at Q4_K_M	Recommended VRAM	Smallest card in our guide that holds it
1.3B	1 GB	2 GB	NVIDIA Jetson Orin Nano 8GB
6.7B	5 GB	8 GB	NVIDIA Jetson Orin Nano 8GB
33B	20 GB	30 GB	NVIDIA RTX 3090

Hardware that runs DeepSeek Coder

Every one of the 19 cards with dedicated memory in our hardware guide holds at least one size of DeepSeek Coder at Q4_K_M. The cheapest that does is the NVIDIA Jetson Orin Nano 8GB at about $199, running the 6.7B.

The largest size any of them holds is the 33B at 20 GB, on the NVIDIA RTX 3090 (24 GB at 936 GB/s). Bandwidth, not capacity, sets the speed once a size fits, so a card further down the list with faster memory will generate more tokens per second on the same weights than a larger, slower one.

What people use it for

The catalogue lists 2 recorded uses for DeepSeek Coder: code completion, VS Code extensions. It is tagged code, completion.

Where DeepSeek Coder sits in the DeepSeek family

Our catalogue holds 4 DeepSeek entries. Ordered by release date, DeepSeek Coder (2023-11) is the 1st of them, and DeepSeek R1 (2025-01) is newer. DeepSeek R1 carries the same recorded memory footprint, so choosing between it and this one is a question of behaviour rather than of hardware: DeepSeek Coder is the one recorded here for small-footprint code completion.

It ships under the DeepSeek License, which is neither a standard open-source licence nor a closed one. Read the terms before shipping anything commercial on top of it.

Context window: 16K

16K tokens comes to roughly 12,000 words at the usual three-quarters-of-a-word-per-token ratio: a long chat, a single source file, one chapter. It is the band where a retrieval step stops being optional, because the interesting documents do not fit. The recommended VRAM figure above assumes a working context rather than the maximum, and filling this one will push past it.

How to run DeepSeek Coder locally

Our catalogue does not record registry tags, so look the current tag up in the Ollama library or on Hugging Face before pasting these. The commands below are the shape of the workflow, not a copy-and-paste recipe.

Option 1: Ollama (simplest)

# find the tag for DeepSeek Coder at ollama.com/library
ollama run <tag>

Option 2: Mullama (production)

mullama pull <tag>
mullama run <tag>

Option 3: llama.cpp (CLI)

# download a GGUF from Hugging Face, searching for "DeepSeek Coder GGUF"
./llama-cli -m model.Q4_K_M.gguf -p "Hello, AI!"

Option 4: Python with Mullama or llama-cpp-python

from mullama import Model, Context
model = Model.load("model.Q4_K_M.gguf", n_gpu_layers=99)
ctx = Context(model, n_ctx=4096)
print(ctx.generate("Hello, AI!", 256))

Sources

Model record from src/data/models.json (2026-06-29); card capacities from src/data/gpus.json (2026-06-29). Upstream: Curated from ollama.com/library + community benchmarks + paperswithcode + huggingface.