OPEN
Valen decision model
Valen is an open decision model family built on Qwen3.5 in 0.8B, 2B and 4B sizes. It reads text, images and video and has a free demo and API.
By Valen Team · First Hugging Face weights upload: 5 Oct 2026
Open Qwen3.5-based decision models for text, images and video. The 1.5 family adds 9B and 27B sizes alongside 0.8B, 2B and 4B. Both Jev System One and OpenAI Decisions request formats are supported. Native 1.5 weights were uploaded on 9 November; the original family dates to October.
Quick answers
What is Valen by Valen Team?
Valen: Open Qwen3.5-based decision models for text, images and video.
Who makes Valen?
Facts
- Sizes
- 1.5: 0.8B, 2B, 4B, 9B and 27B; native, ONNX and GGUF formats
- Licence
- Apache-2.0; licence file included
- Inputs
- Text, images and video, with backend-specific limits
- Runtime
- Custom Valen plugins for native GPU serving; ONNX SDKs; custom GGUF runtime. Official unmodified llama.cpp/Ollama support is not claimed. Prebuilt GGUF 0.1.0 lacks direct video.
- Training data
- The 1.5 training data is not public; the earlier general-training dataset does not establish 1.5 data availability
- Demo
- The free API serves 1.5 in 0.8B, 2B and 4B sizes, according to the current card
Claims and sources
- The maker reports public Decision Index text/vision scores of 69.01/77.85 for 1.5-4B and 74.66/82.91 for 1.5-27B. These are normalized public-track scores, not accuracy percentages or sealed-board results. The published result package records per-size runs. vendor reportedsource
Links
Source: Site and demo → Source: Hugging Face → Source: 1.5 card and weights → Source: 1.5 public result records →
Related projects
No related projects are listed in the registry.