ClawForge - Workflow Completion: leaderboard
Metric: Strict full-pass accuracy (%) on the 80 tasks whose primary ability is workflow completion (closing incomplete multi-step workflows), on the ClawForge-Bench snapshot (17 generator-backed scenarios, 362 executable state-conflict workflow tasks over tasks, calendar, email, messaging, files and runtime configuration), command-line agent in multi mode with full interaction history and a 25-step budget, scored by result-first validators over the final normalized state; higher is better. Source: arxiv.org. Saturation forecast: Not forecast. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.3 Codex | 42.5 |
| 2 | Kimi K2.5 | 37.5 |
| 3 | O3 | 36.2 |
| 4 | GPT-5.4 | 35 |
| 5 | GPT-5.2 | 35 |
| 6 | Claude Opus 4.6 | 32.5 |
| 7 | Claude Sonnet 4.6 | 28.8 |
Interactive version: theaggregate.ai/benchmark?slug=clawforge-workflow-completion · How It Works · Data refreshed daily, snapshot 2026-10-07.