# Remote Labor Index (RLI)

Tests whether AI can complete real paid freelance projects well enough that a client would accept the work.

Category: agents. Type: leaderboard. Status: active.
Metric: Automation rate: share of projects completed acceptably (%) (higher is better).
Source: Scale AI / Center for AI Safety — https://labs.scale.com/leaderboard/rli, retrieved 2026-11-17, license: not stated.

| Rank | Model | Automation rate (%) |
|---|---|---|
| 1 | Fable 5.2 (NEW) | 41.25 |
| 2 | Opus 5.5 | 21.25 |
| 3 | GPT 6 Astra | 20.83 |
| 4 | Fable 5.1 | 17.92 |
| 5 | Gemini Argon 4 | 17.08 |
| 6 | Fable 5 | 15.83 |
| 7 | Deepseek V4.1 Flash (NEW) | 9.58 |
| 8 | Opus 4.8 | 8.33 |
| 9 | Codex GPT 5.5 | 6.25 |
| 10 | Gemini 3.8 Flash | 5.83 |
| 11 | Gemini 3.7 Flash | 5.00 |
| 12 | claude-opus-4-6 (CoWork) | 4.17 |
| 13 | claude-opus-4-5-20251101-thinking | 3.75 |
| 14 | Manus_1.6 (Max) | 2.92 |
| 15 | gpt-5.2-2025-12-11 (medium) | 2.50 |
| 15 | Manus 1.0 | 2.50 |
| 15 | Manus 1.5 | 2.50 |
| 18 | claude-4-5-Sonnet | 2.08 |
| 18 | gpt-5.2-2025-12-11 (default) | 2.08 |
| 20 | gpt-5-2025-08-07 | 1.67 |
| 21 | ChatGPT agent | 1.25 |
| 21 | gemini-3-pro-preview | 1.25 |
| 23 | gemini-2.5-pro-preview-06-05 | 0.83 |

Notes: No update date shown; includes entries tagged NEW. Tied ranks as on page.
