DeepSeek-V4.1-Flash-1017-GSQ-RCO-GGUF
Non-uniform GGUF quantizations of a 763B-parameter, 384-expert MoE, produced with GSQ and RCO.
Params
763.2 B
Context
—
Downloads 30d
3.3 M
Likes
646
Download history
daily snapshots · 16 days3.3 M412 K
Oct 26Oct 31Nov 5Nov 10
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 |
|---|---|---|---|---|
| mmproj-DeepSeek-V4.1-Flash-1017-BF16.gguf | GGUF | 1.0 GB | 116.1 GB | ❌ Won’t fit |
| DeepSeek-V4.1-Flash-1017-GSQ-RCO-IQ2_XS-00002-of-00006.gguf | IQ2_XS | 39.6 GB | 158.5 GB | ❌ Won’t fit |
| DeepSeek-V4.1-Flash-1017-GSQ-RCO-IQ1_M-00004-of-00006.gguf | IQ1_M | 39.7 GB | 158.7 GB | ❌ Won’t fit |
| DeepSeek-V4.1-Flash-1017-GSQ-RCO-IQ2_XXS-00003-of-00006.gguf | IQ2_XXS | 39.7 GB | 158.7 GB | ❌ Won’t fit |
| DeepSeek-V4.1-Flash-Engram-IQ4_NL.gguf | IQ4_NL | 110.6 GB | 236.6 GB | ❌ Won’t fit |
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$ ollama run deepseek-v4-1-flash-1017-gsq-rco-gguf # pin the quantization explicitly $ ollama run deepseek-v4-1-flash-1017-gsq-rco-gguf-iq1_m
$ huggingface-cli download ISTA-DASLab/DeepSeek-V4.1-Flash-1017-GSQ-RCO-GGUF-GGUF \
DeepSeek-V4.1-Flash-1017-GSQ-RCO-IQ1_M-00004-of-00006.gguf --local-dir .
$ llama-cli -m DeepSeek-V4.1-Flash-1017-GSQ-RCO-IQ1_M-00004-of-00006.gguf \
-c 8192 -ngl 99 -t 8 --color
$ curl -s https://aimodelscomparison.com/api/v1/models/deepseek-v4-1-flash-1017-gsq-rco-gguf
{
"hf_id": "ISTA-DASLab/DeepSeek-V4.1-Flash-1017-GSQ-RCO-GGUF",
"params_b": 763.21,
"context_length": null,
"license": { "id": "mit", "commercial": "yes" },
"downloads_30d": 3270121,
"vram_estimates": [
{ "quant": "IQ1_M", "gb": 158.7 },
{ "quant": "IQ2_XS", "gb": 158.5 }
],
"updated_at": "2026-10-29T01:00:05Z"
}
Specifications
- Parameters
- 763.2 B
- Licence
- mit
- First seen on the Hub
- 2026-10-23
- Base model
- DeepSeek-V4.1-Flash-1017
- Training datasets
- undisclosed
- Added to our catalog
- 2026-10-26
Family
Base model and the most-downloaded derivatives in the catalog.
Compare with any text-generation model
Popular comparisons