Best open vision-language models
Vision-language models process both images and text. They can caption images, answer questions about images, and chat about visual content. These models combine computer vision and language understanding.
Last updated 3 November 2026 · refreshed after every Hugging Face sync · 15 models
| # | Model | Size | Memory | License | Downloads (30 days) | Likes |
|---|---|---|---|---|---|---|
| 1 | Qwen3.8-27B Qwen | 27B · FP16 | ~65 GB | Apache 2.0 | 7.9M | 19K |
| 2 | gemma-4-31B-it | 31B · FP16 | ~75 GB | Apache 2.0 | 8.1M | 4.3K |
| 3 | GLM-5.3-Flash zai-org | – · FP16 | – | MIT | 6.3M | 3K |
| 4 | Qwen3.8-Flash-Next Qwen | – · FP16 | – | Custom (other) | 3.9M | 6.7K |
| 5 | Qwen3.5-9B Qwen | 9B · FP16 | ~22 GB | Apache 2.0 | 6.4M | 2.3K |
| 6 | Unlimited-OCR baidu | – · FP16 | – | MIT | 4.2M | 4.9K |
| 7 | Qwen3.6-35B-A3B Qwen | 35B · FP16 | ~84 GB | Apache 2.0 | 4.6M | 3.1K |
| 8 | Qwen3.8-Flash-Next-GSQ-RCO-GGUF ISTA-DASLab | – · GGUF | – | Apache 2.0 | 7.4M | 1.1K |
| 9 | gemma-4-26B-A4B-it | 26B · FP16 | ~63 GB | Apache 2.0 | 5.9M | 1.7K |
| 10 | Qwen2.5-VL-7B-Instruct Qwen | 7B · FP16 | ~17 GB | Apache 2.0 | 5.9M | 1.8K |
| 11 | Qwen3-VL-8B-Instruct Qwen | 8B · FP16 | ~20 GB | Apache 2.0 | 6.9M | 1.2K |
| 12 | Qwen3.5-4B Qwen | 4B · FP16 | ~10 GB | Apache 2.0 | 6.4M | 1.2K |
| 13 | DeepSeek-V4.1-Flash deepseek-ai | – · FP16 | – | MIT | 2.1M | 4.8K |
| 14 | Qwen3.8-27B-FP8 Qwen | 27B · FP16 | ~65 GB | Apache 2.0 | 4.7M | 955 |
| 15 | JEV-27B-VL autotrust | 27B · FP16 | ~65 GB | Apache 2.0 | 2.5M | 3.1K |
How we rank
Models with the image-text-to-text pipeline tag. We include models seen on Hugging Face in the last 14 days with at least 10,000 downloads in the last 30 days. Score = log10(30-day downloads) + ½ × log10(likes + 1), so popularity leads and likes break near-ties. We keep one entry per model name and leave out adult and uncensored builds. Memory is estimated from parameter count and the published file format (16-bit ≈ 2 bytes per parameter, 4-bit ≈ 0.5) plus 20% headroom; long contexts need more. A quantised copy of a bigger model may also fit — check the model page.