ByteDance's WorkflowGym Exposes GUI Agents: Top Models Pass Only 30%
机器之心 · wechat · 2026-07-23
ByteDance's Seed evaluation team has released WorkflowGym, a new benchmark that tests GUI agents on professional software like FreeCAD and Blender, revealing the true boundaries of current model capabilities.
Evaluation Design & Results:
- The benchmark covers 56 professional applications and 338 real-world, long-horizon workflows, averaging over 100 operational steps.
- In these rigorous tasks, top-tier models like Gemini 3.1 Pro and Kimi K2.6 achieved a single-pass success rate of just over 30%. Lighter models almost completely failed on tasks requiring more than 60 steps.
Core Model Flaws:
The paper identifies four major dead spots in current agent frameworks: loss of long-context consistency, cascading errors without self-correction, goal drift, and a lack of understanding of professional UI paradigms. This suggests that scaling parameters alone won't fix these architectural bottlenecks, making human-AI collaboration the most practical product approach for now.
More from Models
- Moonshot’s reported Anthropic distillation draws terms-of-service and strategy claims — connoraxiotes · 2026-07-23
- Open-weight Chinese models are still not the same as permissive open source — tamla_htoo · 2026-07-23
- Chinese open-source models may be reinforcing NVIDIA’s inference moat — zephyr_z9 · 2026-07-23
- Kimi generates luxury watch renders with a detailed mechanical movement — cedric_chee · 2026-07-23
- Qwen3.8-Max looks stronger on 3D web frontends than Kimi K3 — cedric_chee · 2026-07-23
- Poster says Opus-5 may beat Fable on cost and performance — adonis_singh · 2026-07-23