# General AI Assistant Tasks on GAIA Level 2 original (test)

Source: [General AI Assistant Tasks on GAIA Level 2 original (test)](<https://www.sota2.com/research/sota/general-ai-assistant-tasks-on-gaia-level-2-original-test>)

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

Data updated: 2026-02-26T05:57:47.369Z

Evaluation of general AI assistant capabilities on the original test set specifically for GAIA Level 2, measured by performance percentage, cost, and delay.

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 | General AI Assistant Tasks |
| Dataset (including any stated split) | GAIA Level 2 original (test) |
| Metric | Perf (%) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 15 of 15 |
| 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

- [Perf (%)](<https://www.sota2.com/research/sota/general-ai-assistant-tasks-on-gaia-level-2-original-test.md?metric=Perf%20(%25)>): Higher is better; unit: PERCENT.

- [Cost ($)](<https://www.sota2.com/research/sota/general-ai-assistant-tasks-on-gaia-level-2-original-test.md?metric=Cost%20(%24)>): Lower is better; unit: OTHER.

- [Delay (h)](<https://www.sota2.com/research/sota/general-ai-assistant-tasks-on-gaia-level-2-original-test.md?metric=Delay%20(h)>): Lower is better; unit: TIME.

## Recorded results

| Method | Perf (%) | Result date | Report paper |
| --- | --- | --- | --- |
| EvoRoute | 59.3 | Not recorded | 695de87753fdd5c361743c7c |
| Claude-4 | 58.14 | Not recorded | 695de87753fdd5c361743c7c |
| EvoRoute | 53.73 | Not recorded | 695de87753fdd5c361743c7c |
| MasRouter | 49.74 | Not recorded | 695de87753fdd5c361743c7c |
| GPT-4.1 | 49.73 | Not recorded | 695de87753fdd5c361743c7c |
| PromptLLM | 47.65 | Not recorded | 695de87753fdd5c361743c7c |
| GraphRouter | 46.23 | Not recorded | 695de87753fdd5c361743c7c |
| Claude-4 | 46.23 | Not recorded | 695de87753fdd5c361743c7c |
| MasRouter | 42.07 | Not recorded | 695de87753fdd5c361743c7c |
| PromptLLM | 40.92 | Not recorded | 695de87753fdd5c361743c7c |
| GraphRouter | 39.78 | Not recorded | 695de87753fdd5c361743c7c |
| GPT-4o | 39.22 | Not recorded | 695de87753fdd5c361743c7c |
| GPT-4.1 | 36.3 | Not recorded | 695de87753fdd5c361743c7c |
| Qwen3-14b | 10.34 | Not recorded | 695de87753fdd5c361743c7c |
| Qwen3-14b | 4.73 | Not recorded | 695de87753fdd5c361743c7c |

### EvoRoute

Perf (%): 59.3. Result ID: 6996a2e35698665ec4ec70c1.

[Reporting paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

Recorded evaluation configuration: {"Framework":"CK-Pro","Setting":"Ours"}. This record does not assert independent reproduction.

### Claude-4

Perf (%): 58.14. Result ID: 6996a2e35698665ec4ec70bb.

[Reporting paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

Recorded evaluation configuration: {"Framework":"CK-Pro","Setting":"Manual"}. This record does not assert independent reproduction.

### EvoRoute

Perf (%): 53.73. Result ID: 6996a2e35698665ec4ec70c8.

[Reporting paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

Recorded evaluation configuration: {"Framework":"Smolagent","Setting":"Ours"}. This record does not assert independent reproduction.

### MasRouter

Perf (%): 49.74. Result ID: 6996a2e35698665ec4ec70c0.

[Reporting paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76fd1e209131cea7049b.md>)

Recorded evaluation configuration: {"Framework":"CK-Pro","Setting":"Routing"}. This record does not assert independent reproduction.

### GPT-4.1

Perf (%): 49.73. Result ID: 6996a2e35698665ec4ec70bd.

[Reporting paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

Recorded evaluation configuration: {"Framework":"CK-Pro","Setting":"Manual"}. This record does not assert independent reproduction.

### PromptLLM

Perf (%): 47.65. Result ID: 6996a2e35698665ec4ec70be.

[Reporting paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa777b1e209131cea737d1.md>)

Recorded evaluation configuration: {"Framework":"CK-Pro","Setting":"Routing"}. This record does not assert independent reproduction.

### GraphRouter

Perf (%): 46.23. Result ID: 6996a2e35698665ec4ec70bf.

[Reporting paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa777b1e209131cea737d1.md>)

Recorded evaluation configuration: {"Framework":"CK-Pro","Setting":"Routing"}. This record does not assert independent reproduction.

### Claude-4

Perf (%): 46.23. Result ID: 6996a2e35698665ec4ec70c3.

[Reporting paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

Recorded evaluation configuration: {"Framework":"Smolagent","Setting":"Manual"}. This record does not assert independent reproduction.

### MasRouter

Perf (%): 42.07. Result ID: 6996a2e35698665ec4ec70c7.

[Reporting paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76fd1e209131cea7049b.md>)

Recorded evaluation configuration: {"Framework":"Smolagent","Setting":"Routing"}. This record does not assert independent reproduction.

### PromptLLM

Perf (%): 40.92. Result ID: 6996a2e35698665ec4ec70c5.

[Reporting paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa777b1e209131cea737d1.md>)

Recorded evaluation configuration: {"Framework":"Smolagent","Setting":"Routing"}. This record does not assert independent reproduction.

### GraphRouter

Perf (%): 39.78. Result ID: 6996a2e35698665ec4ec70c6.

[Reporting paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa777b1e209131cea737d1.md>)

Recorded evaluation configuration: {"Framework":"Smolagent","Setting":"Routing"}. This record does not assert independent reproduction.

### GPT-4o

Perf (%): 39.22. Result ID: 6996a2e35698665ec4ec70bc.

[Reporting paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

Recorded evaluation configuration: {"Framework":"CK-Pro","Setting":"Manual"}. This record does not assert independent reproduction.

### GPT-4.1

Perf (%): 36.3. Result ID: 6996a2e35698665ec4ec70c4.

[Reporting paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

Recorded evaluation configuration: {"Framework":"Smolagent","Setting":"Manual"}. This record does not assert independent reproduction.

### Qwen3-14b

Perf (%): 10.34. Result ID: 6996a2e35698665ec4ec70ba.

[Reporting paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

Recorded evaluation configuration: {"Framework":"CK-Pro","Setting":"Manual"}. This record does not assert independent reproduction.

### Qwen3-14b

Perf (%): 4.73. Result ID: 6996a2e35698665ec4ec70c2.

[Reporting paper](<https://www.sota2.com/research/paper/695de87753fdd5c361743c7c.md>)

Recorded evaluation configuration: {"Framework":"Smolagent","Setting":"Manual"}. This record does not assert independent reproduction.

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