Agents' Last Exam v1.1 - Claude Code Partial Credit: leaderboard
ALE v1.1, 122 tasks. Native harness/model/effort roster without selecting strongest effort. Pass rate counts perfect-score runs; partial-credit score averages graded scores over all 122 tasks, including zero for failed/missing runs. Timeouts retain graded scores. Separate from legacy ALE v1. Claude Code; Partial Credit. Metric: Partial Credit (%).
Metric: Partial Credit (%). Source: agents-last-exam.org. Saturation forecast: Rough model projection: around 2026. 2 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | claude-opus-5-5 (ALE Run: harness=Claude%20Code; effort=max; source-model=claude-opus-5-5) | 67.48 | |
| 2 | minimax-m3-1-flash (ALE Run: harness=Claude%20Code; effort=max; source-model=minimax-m3-1-flash) | 57.92 |
Interactive version: theaggregate.ai/benchmark?slug=agents-last-exam-v1-1-claude-code-partial-credit · How It Works · Data refreshed daily, snapshot 2026-10-27.