# Long-horizon agentic tasks on HLE Full

Source: [Long-horizon agentic tasks on HLE Full](<https://www.sota2.com/research/sota/long-horizon-agentic-tasks-on-hle-full>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-31T09:42:17.340Z

Evaluation of model performance on long-horizon agentic tasks using the HLE Full dataset, measured by Pass@1.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Long-horizon agentic tasks |
| Dataset (including any stated split) | HLE Full |
| Metric | Pass@1 |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 10 of 10 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Pass@1](<https://www.sota2.com/research/sota/long-horizon-agentic-tasks-on-hle-full.md?metric=Pass%401>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Pass@1 | Result date | Report paper |
| --- | --- | --- | --- |
| Gemini-3.0-Pro | 45.8 | Not recorded | 69cb555c0636e5bf4afca207 |
| Claude-4.5-Opus | 43.4 | Not recorded | 69cb555c0636e5bf4afca207 |
| GPT-5.1 High | 42.7 | Not recorded | 69cb555c0636e5bf4afca207 |
| DeepSeek-v3.2 | 40.8 | Not recorded | 69cb555c0636e5bf4afca207 |
| Tongyi-DR-30B-A3B | 32.9 | Not recorded | 69cb555c0636e5bf4afca207 |
| AgentFounder-30B-A3B | 31.5 | Not recorded | 69cb555c0636e5bf4afca207 |
| MiroThinker-v1.5-30B-A3B | 31 | Not recorded | 69cb555c0636e5bf4afca207 |
| IterResearch-30B-A3B | 28.8 | Not recorded | 69cb555c0636e5bf4afca207 |
| OpenAI DeepResearch | 26.6 | Not recorded | 69cb555c0636e5bf4afca207 |
| ASearcher-Web-32B | 12.5 | Not recorded | 69cb555c0636e5bf4afca207 |

### Gemini-3.0-Pro

Pass@1: 45.8. Result ID: 69cb96f447f34ce924b4432d.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555c0636e5bf4afca207.md>)

Recorded evaluation configuration: {"Context Management":"w/o CM"}. This record does not assert independent reproduction.

### Claude-4.5-Opus

Pass@1: 43.4. Result ID: 69cb96f447f34ce924b4432c.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555c0636e5bf4afca207.md>)

Recorded evaluation configuration: {"Context Management":"w/o CM"}. This record does not assert independent reproduction.

### GPT-5.1 High

Pass@1: 42.7. Result ID: 69cb96f447f34ce924b4432e.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555c0636e5bf4afca207.md>)

Recorded evaluation configuration: {"Context Management":"w/o CM"}. This record does not assert independent reproduction.

### DeepSeek-v3.2

Pass@1: 40.8. Result ID: 69cb96f447f34ce924b44334.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555c0636e5bf4afca207.md>)

[Method origin paper](<https://www.sota2.com/research/paper/692fc410d6b0dd33429332f4.md>)

Recorded evaluation configuration: {"Context Management":"Baseline (w/o CM)"}. This record does not assert independent reproduction.

### Tongyi-DR-30B-A3B

Pass@1: 32.9. Result ID: 69cb96f447f34ce924b44335.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555c0636e5bf4afca207.md>)

Recorded evaluation configuration: {"Context Management":"Baseline (w/o CM)"}. This record does not assert independent reproduction.

### AgentFounder-30B-A3B

Pass@1: 31.5. Result ID: 69cb96f447f34ce924b44332.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555c0636e5bf4afca207.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b671e209131cea8833d.md>)

Recorded evaluation configuration: {"Context Management":"-"}. This record does not assert independent reproduction.

### MiroThinker-v1.5-30B-A3B

Pass@1: 31. Result ID: 69cb96f447f34ce924b44333.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555c0636e5bf4afca207.md>)

Recorded evaluation configuration: {"Context Management":"-"}. This record does not assert independent reproduction.

### IterResearch-30B-A3B

Pass@1: 28.8. Result ID: 69cb96f447f34ce924b44331.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555c0636e5bf4afca207.md>)

Recorded evaluation configuration: {"Context Management":"-"}. This record does not assert independent reproduction.

### OpenAI DeepResearch

Pass@1: 26.6. Result ID: 69cb96f447f34ce924b4432f.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555c0636e5bf4afca207.md>)

Recorded evaluation configuration: {"Context Management":"-"}. This record does not assert independent reproduction.

### ASearcher-Web-32B

Pass@1: 12.5. Result ID: 69cb96f447f34ce924b44330.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555c0636e5bf4afca207.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b441e209131cea87635.md>)

Recorded evaluation configuration: {"Context Management":"-"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.