# Assistant Tasks on GAIA (official)

Source: [Assistant Tasks on GAIA (official)](<https://www.sota2.com/research/sota/assistant-tasks-on-gaia-official>)

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

Data updated: 2026-02-26T05:42:28.131Z

Evaluation of assistant capabilities on the official split of the GAIA dataset, measured by Exact Match percentage.

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 | Assistant Tasks |
| Dataset (including any stated split) | GAIA (official) |
| Metric | Exact Match |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 6 of 6 |
| 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

- [Exact Match](<https://www.sota2.com/research/sota/assistant-tasks-on-gaia-official.md?metric=Exact%20Match>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Exact Match | Result date | Report paper |
| --- | --- | --- | --- |
| Tendem’s AI agent | 78.2 | Not recorded | 69818138a8b01e76cc8e238f |
| Flowith | 78.1 | Not recorded | 69818138a8b01e76cc8e238f |
| Manus | 73.4 | Not recorded | 69818138a8b01e76cc8e238f |
| HAL Generalist Agent | 72.6 | Not recorded | 69818138a8b01e76cc8e238f |
| ChatGPT Deep Research | 67.4 | Not recorded | 69818138a8b01e76cc8e238f |
| HF Open Deep Research | 62.8 | Not recorded | 69818138a8b01e76cc8e238f |

### Tendem’s AI agent

Exact Match: 78.2. Result ID: 699658345698665ec49d8cd1.

[Reporting paper](<https://www.sota2.com/research/paper/69818138a8b01e76cc8e238f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69818138a8b01e76cc8e238f.md>)

Recorded evaluation configuration: {"human involvement":"none"}. This record does not assert independent reproduction.

### Flowith

Exact Match: 78.1. Result ID: 699658345698665ec49d8cd4.

[Reporting paper](<https://www.sota2.com/research/paper/69818138a8b01e76cc8e238f.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Manus

Exact Match: 73.4. Result ID: 699658345698665ec49d8cd3.

[Reporting paper](<https://www.sota2.com/research/paper/69818138a8b01e76cc8e238f.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### HAL Generalist Agent

Exact Match: 72.6. Result ID: 699658345698665ec49d8cd6.

[Reporting paper](<https://www.sota2.com/research/paper/69818138a8b01e76cc8e238f.md>)

Recorded evaluation configuration: {"backbone":"Claude Sonnet 4.5"}. This record does not assert independent reproduction.

### ChatGPT Deep Research

Exact Match: 67.4. Result ID: 699658345698665ec49d8cd2.

[Reporting paper](<https://www.sota2.com/research/paper/69818138a8b01e76cc8e238f.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### HF Open Deep Research

Exact Match: 62.8. Result ID: 699658345698665ec49d8cd5.

[Reporting paper](<https://www.sota2.com/research/paper/69818138a8b01e76cc8e238f.md>)

Recorded evaluation configuration: {"model":"GPT-5 medium"}. This record does not assert independent reproduction.

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