beipo-ai/NorthCUA-Pro
just landed · new, unvettedmodel
New, unvetted: created or first seen in the last few days, not from a trusted org and not past the signal threshold yet.
NorthCUA-Pro is a 397B-parameter computer-use model with 17B active parameters, built on Qwen3.5. Mixture-of-experts, 397B total, 17B active (512 experts, 10 per token). Context: 256K tokens (config). License: apache-2.0.
- Org
- beipo-ai
- Released
- 2026-10-30
- Status
- just landed · new, unvetted
- License
- apache-2.0 (permissive)
- Parameters
- 396.8B total, 17.0B active
- Architecture
- MoE · 512 experts, 10/token · Qwen3_5MoeForConditionalGeneration
- Context
- 262,144 tokens
- Task
- image-text-to-text · vision
- Signal
- 5 HF likes · 361 downloads
Runs on
mixed: real files + estimate; need = weights + FP16 KV cache for 8K tokens + 1GB. A sizing calculation, not tested.
- Q4 ~240.1 GB (estimated size): Mac 512GB
- Q8 ~423.6 GB (estimated size): no listed device
- FP16 ~795.6 GB: no listed device
Links
First published here 2026-10-30 · updated 2026-11-02 · markdown · JSON