Qwen beats ChatGPT at his tasks, now he hunts a UI to run it on whole code projects
yeah280 · reddit · 2026-09-09
A paying ChatGPT user reports that Qwen often understands his intent better than ChatGPT/GPT-5.6, and is considering dropping his subscription to make Qwen his primary model (possibly adding Claude later).
What he's missing is the workspace around the model: upload a full ZIP/code project, let the model inspect the structure, iteratively edit multiple files, and get the modified project back — with a proper UI instead of raw API calls. He already has an OpenRouter key and asks which desktop app, web UI, IDE integration or open-source coding agent supports this, and whether paying for Qwen (vs. the free website) buys better models, higher limits or real project-level coding features, or is essentially just API usage.
More from coding & agent
- ExecCritic: test-verify-revise RL scaffold lifts Qwen3.5 on SWE-bench from 61.2% to 72.6% — SharonYixuanLi · 2026-09-09
- Claude Code user burns $4,500 of API spend in 30 days on a $200/month plan — PratikKadam_ · 2026-09-09
- Claude Desktop Showed 4.7k Context for Our 408-Tool MCP Server — That Number Is a Trap — dennyriedl · 2026-09-09
- TesterArmy runs AI QA agents on every PR, claims 150 hours saved monthly per team — marclou · 2026-09-09
- This IDE Works Like a City Builder: Open-Source 2D Canvas for Managing AI Agents — thisiskp_ · 2026-09-09
- Demo: Combining Opus 5, Fable 5.1 and Blender into One Creative Workflow — prasenx · 2026-09-09