DeepSeek-V4.1-Flash-1017-Maya-GGUF
GGUF quantizations of deepseek-ai/DeepSeek-V4.1-Flash-1017 (a mixture of experts: 384 routed experts per layer, 6 active per token, plus a shared expert), made for Project Maya. Project Maya keeps the most-used experts on the GPU and reads the rest from RAM and the SSD, so the model runs well below its size. Both quants are made from DeepSeek's own release (FP4 experts, FP8 dense weights), not from a re-quantized fil...
Params
763.2 B
Context
—
Downloads 30d
207 K
Likes
65
Download history
daily snapshots · 10 days207 K147 K
Nov 1Nov 4Nov 7Nov 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 |
|---|---|---|---|---|
| DeepSeek-V4.1-Flash-1017-Maya-S-IQ2_XXS-00006-of-00006.gguf | IQ2_XXS | 80.6 GB | 203.7 GB | ❌ Won’t fit |
| DeepSeek-V4.1-Flash-1017-Maya-XS-IQ1_S-00005-of-00005.gguf | IQ1_S | 80.6 GB | 203.7 GB | ❌ Won’t fit |
| model.safetensors | fp16 | 203.1 GB | 338.4 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-maya-gguf # pin the quantization explicitly $ ollama run deepseek-v4-1-flash-1017-maya-gguf-iq2_xxs
$ huggingface-cli download peasantsmith/DeepSeek-V4.1-Flash-1017-Maya-GGUF-GGUF \
DeepSeek-V4.1-Flash-1017-Maya-S-IQ2_XXS-00006-of-00006.gguf --local-dir .
$ llama-cli -m DeepSeek-V4.1-Flash-1017-Maya-S-IQ2_XXS-00006-of-00006.gguf \
-c 8192 -ngl 99 -t 8 --color
$ curl -s https://aimodelscomparison.com/api/v1/models/deepseek-v4-1-flash-1017-maya-gguf
{
"hf_id": "peasantsmith/DeepSeek-V4.1-Flash-1017-Maya-GGUF",
"params_b": 763.21,
"context_length": null,
"license": { "id": "mit", "commercial": "yes" },
"downloads_30d": 207147,
"vram_estimates": [
{ "quant": "IQ2_XXS", "gb": 203.7 },
{ "quant": "IQ1_S", "gb": 203.7 }
],
"updated_at": "2026-11-02T01:00:24Z"
}
Specifications
- Parameters
- 763.2 B
- Licence
- mit
- First seen on the Hub
- 2026-10-19
- Base model
- DeepSeek-V4.1-Flash-1017
- Training datasets
- undisclosed
- Added to our catalog
- 2026-11-01
Family
Base model and the most-downloaded derivatives in the catalog.
Compare with any text-generation model