MODEL COMPARISON

Tradeoffs, side by side.

Select up to four models. These signals narrow the shortlist; benchmark on your own workload before deploying.

Signalunsloth/Qwen3-VL-32B-Thinking-GGUF
Taskimage-text-to-text
Parameters32.0B
Minimum VRAM24 GB
Recommended VRAM48 GB
Licenseapache-2.0
Commercial signalLIKELY_ALLOWED
Downloads2,945
Likes26
SourceHugging Face ↗