# Agents' Last Exam (Linux) leaderboard

Agentic benchmark

> Every figure below is reproduced as its upstream source published it: nothing is modelled, estimated, interpolated or converted. `Unavailable` means no source published the value — it is never a zero. Each value carries its evidence state and the date it was observed.

Page: https://modelscale.dev/leaderboards/benchmarks/agentic-agents-last-exam-linux

Agents' Last Exam, Linux subset. Every model the catalog carries a published Agents' Last Exam (Linux) value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Agentic |
| Measure | Provider-reported task score |
| Tasks | Linux agent tasks (subset size not published) |
| Difficulty | Advanced agentic work |
| Published by | [Introducing Muse Ember](https://research.meta.ai/blog/muse-ember) |

The Linux subset of Agents' Last Exam, an end-to-end agent task benchmark, as reported by Meta.

## Agents' Last Exam (Linux) ranking

3 models with a published Agents' Last Exam (Linux) value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.

| Rank | Model | Provider | Provider-reported task score | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [Muse Ember](https://modelscale.dev/models/muse-ember) | Meta | 61.6 | Observed 2026-11-21 · source benchlm:models |
| 2 | [GPT-6 Luna](https://modelscale.dev/models/gpt-6-luna) | OpenAI | 57.1 | Observed 2026-11-21 · source benchlm:models |
| 3 | [Claude Haiku 5.5](https://modelscale.dev/models/claude-haiku-5-5) | Anthropic | 57.0 | Observed 2026-11-21 · source benchlm:models |

Rows are ordered by the value the source published, highest first. The source does not state whether a higher value is the better result, so this document does not either: for a benchmark that measures a rate of failure, read the table from the bottom.
