POCKET-35B-GGUF
> ### 📚 Collections > ▶ POCKET Models — this family (on-device, no GPU) > Darwin Family · Aether Foundation · VKAE Accelerated
Params
—
Context
—
Downloads 30d
812 K
Likes
72
Download history
daily snapshots · 39 days
▲ 461 K in the last 30 days (131.2%)
826 K379 K
Aug 23Sep 2Sep 12Sep 21
826 K189 K
Aug 14Aug 27Sep 9Sep 21
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 |
|---|---|---|---|---|
| POCKET-35B-IQ1_M.gguf | IQ1_M | 8.2 GB | 9.6 GB | ✅ Runs comfortably |
| POCKET-35B-Q2_K.gguf | Q2_K | 12.9 GB | 14.7 GB | ✅ Runs comfortably |
| POCKET-35B-Q3_K_M.gguf | Q3_K_M | 16.8 GB | 18.9 GB | ✅ Runs comfortably |
| POCKET-35B-Q4_K_M.gguf | Q4_K_M | 21.2 GB | 23.8 GB | ⚠️ Tight — reduce context |
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 pocket-35b-gguf # pin the quantization explicitly $ ollama run pocket-35b-gguf-q4_k_m
$ huggingface-cli download FINAL-Bench/POCKET-35B-GGUF-GGUF \
POCKET-35B-Q4_K_M.gguf --local-dir .
$ llama-cli -m POCKET-35B-Q4_K_M.gguf \
-c 8192 -ngl 99 -t 8 --color
$ curl -s https://aimodelscomparison.com/api/v1/models/pocket-35b-gguf
{
"hf_id": "FINAL-Bench/POCKET-35B-GGUF",
"params_b": null,
"context_length": null,
"license": { "id": "apache-2.0", "commercial": "yes" },
"downloads_30d": 811530,
"vram_estimates": [
{ "quant": "IQ1_M", "gb": 9.6 },
{ "quant": "Q2_K", "gb": 14.7 }
],
"updated_at": "2026-08-14T01:00:34Z"
}
Specifications
- Licence
- apache-2.0
- First seen on the Hub
- 2026-07-22
- Training datasets
- undisclosed
- Added to our catalog
- 2026-08-14
Compare with any text-generation model