Open Source Qwen3.8-VL Universal Prompt Architect Model
Old_Estimate1905 · reddit · 2026-08-16
An Ollama model based on Qwen3.8-27B (Abliterated) designed for generative AI workflows. It features zero filler output, multi-engine dispatching (FLUX.2, MiniMax H3, Ideogram, etc.), vision grounding, and an uncensored reasoning layer. Supports optimized prompt formats for various generators via specific trigger tags.
More from Apps
- When Simple Rules Beat AI: A Frozen Ride-Hailing State Machine at Dulles — shashib · 2026-10-16
- User Receives $193 State Refund Check Thanks to AI Tool Muse — billyjhowell · 2026-10-16
- Beff Jezos: Google Keyboard's On-Device Speech-to-Text Is Low-Key Insane — beffjezos · 2026-10-16
- AI social gaming platform Astrocade hits 10 million monthly active players — andrey_kurenkov · 2026-10-16
- No-dev-experience game built in 6 weeks with Claude orchestrating Unity and asset pipeline — Gambo7592 · 2026-10-16
- 75-80% of my AI requests now go through Dots: a power user's month-long review — nickbaumann_ · 2026-10-16