deepseek-ai / text-generation updated 2 years ago

deepseek-coder-7b-instruct-v1.5

Name: deepseek-coder-7b-instruct-v1.5
Author: deepseek-ai

Deepseek-Coder-7B-Instruct-v1.5 is continue pre-trained from Deepseek-LLM 7B on 2T tokens by employing a window size of 4K and next token prediction objective, and then fine-tuned on 2B tokens of instruction data.

Params

6.9 B

Context

4,096

Downloads 30d

599 K

Likes

158

Licence: other Not gated SAFETENSORS View on Hugging Face ↗

Download history

daily snapshots · 10 days

599 K484 K

Jul 28Jul 31Aug 3Aug 6

Can you run it?

Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.

File	Quant	Size	Est. VRAM	Verdict on RTX 4090 · 24 GB
model.safetensors	bf16	13.8 GB	16.7 GB	✅ Runs comfortably

Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.

Run it

copy-paste, exact tags checked against the Hub

~ · curl · api/v1

$ curl -s https://aimodelscomparison.com/api/v1/models/deepseek-coder-7b-instruct-v1-5
{
  "hf_id": "deepseek-ai/deepseek-coder-7b-instruct-v1.5",
  "params_b": 6.91,
  "context_length": 4096,
  "license": { "id": "other", "commercial": "unknown" },
  "downloads_30d": 598564,
  "vram_estimates": [
    { "quant": "bf16", "gb": 16.7 }
  ],
  "updated_at": "2026-07-28T18:05:26Z"
}

est. VRAM —on RTX 4090 · 24 GBJSON API →

Specifications

Architecture: LlamaForCausalLM
Parameters: 6.9 B
Tensor type: BF16
Context length: 4,096
Vocabulary: 102,400
Layers / heads: 30 / 32
Licence: other
First seen on the Hub: 2024-01-25
Training datasets: undisclosed
Added to our catalog: 2026-07-28

Compare with

Qwen3-8B 16.3 M Qwen2.5-7B-Instruct 12.1 M Qwen3.6-35B-A3B-NVFP4 11.3 M Llama-3.1-8B-Instruct 8.0 M Qwen2.5-7B-Instruct-AWQ 4.6 M granite-4.1-8b 4.2 M

Appears in

Best open coding models Commercial-use text models Models that fit in 12 GB