Qwen / image-text-to-text updated 1 year ago

Qwen2.5-VL-32B-Instruct

Name: Qwen2.5-VL-32B-Instruct
Author: Qwen

In addition to the original formula, we have further enhanced Qwen2.5-VL-32B's mathematical and problem-solving abilities through reinforcement learning. This has also significantly improved the model's subjective user experience, with response styles adjusted to better align with human preferences. Particularly for objective queries such as mathematics, logical reasoning, and knowledge-based Q&A, the level of detail...

Params

33.5 B

Context

128,000

Downloads 30d

1.1 M

Likes

498

Commercial use: allowed apache-2.0 Not gated SAFETENSORS 1 languages View on Hugging Face ↗

Download history

daily snapshots · 10 days

1.1 M852 K

Jul 28Jul 31Aug 3Aug 6

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
model.safetensors	bf16	68.3 GB	80.6 GB	❌ Won’t fit
model.safetensors (bf16, full)	bf16 + 125K ctx	68.3 GB	154.0 GB	❌ Won’t fit

Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.

Specifications

Architecture: Qwen2_5_VLForConditionalGeneration
Parameters: 33.5 B
Tensor type: BF16
Context length: 128,000
Vocabulary: 152,064
Layers / heads: 64 / 40
Licence: apache-2.0
First seen on the Hub: 2025-03-21
Training datasets: undisclosed
Added to our catalog: 2026-07-28

Family

Base model and the most-downloaded derivatives in the catalog.

Derivative Qwen2.5-VL-32B-Instruct-AWQ Qwen 1.4 M

Compare with

gemma-4-31B-it 11.9 M gemma-4-26B-A4B-it 11.7 M Qwen3.6-35B-A3B-FP8 8.9 M Qwen2.5-VL-3B-Instruct 8.2 M Qwen3.6-27B-FP8 7.8 M Qwen3.6-27B 7.0 M

Appears in

Best open coding models Commercial-use text models Models that fit in 12 GB