Qwen2.5-VL-3B-Instruct vs Gemma-4-E4B-Uncensored-HauhauCS-Aggressive
Specs, VRAM requirements and download trends — updated 6 August 2026.
HauhauCS · B
Params
—
Context
—
30d
937 K
gemma licence · conditional
Specification comparison
Differences are highlighted; identical values are muted.
| Specification | Qwen2.5-VL-3B-Instruct | Gemma-4-E4B-Uncensored-HauhauCS-Aggressive |
|---|---|---|
| Parameters | 3.8 B | — |
| Architecture | Qwen2_5_VLForConditionalGeneration | — |
| Context length | 128,000 | — |
| Licence | — | gemma |
| Commercial use | Unknown | Conditional |
| Languages | 1 | 2 |
| Downloads 30d | 8,170,451 | 937,284 |
| Downloads all time | 97.0 M | 3.5 M |
| Quantizations on the Hub | SAFETENSORS | GGUF |
| Gated | No | No |
| First seen on the Hub | 2025-01-26 | 2026-04-02 |
Download trend
Daily snapshots, last 10 days (28 Jul – 6 Aug)
8.2 M887 K
Qwen2.5-VL-3B-Instruct
Gemma-4-E4B-Uncensored-HauhauCS-Aggressive
VRAM side by side
On RTX 4090 · 24 GB · 8K context unless noted
Gemma-4-E4B-Uncensored-HauhauCS-Aggressive · Q4_K_M
6.4 GB / 24 GB
Qwen2.5-VL-3B-Instruct · fp16 @ 125K ctx
17.5 GB / 24 GB