# General AI Assistant Reasoning on GAIA-Text-103 1.0 (test) Source: [General AI Assistant Reasoning on GAIA-Text-103 1.0 (test)]() Collection index: [llms.txt]() Data updated: 2026-04-23T09:52:39.366Z Evaluation of General AI Assistant Reasoning capabilities on the test split of GAIA-Text version 1.0, measured by Level 1, Level 2, Level 3, and Overall accuracy percentages. 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 Reasoning | | Dataset (including any stated split) | GAIA-Text-103 1.0 (test) | | Metric | Overall Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 26 of 26 | | 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 - [Overall Accuracy](): Higher is better; unit: PERCENT. - [L1 Accuracy](): Higher is better; unit: PERCENT. - [L2 Accuracy](): Higher is better; unit: PERCENT. - [L3 Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Overall Accuracy | Result date | Report paper | | --- | --- | --- | --- | | Claude-3.7-Sonnet | 62.1 | Not recorded | 6982d2aba8b01e76cc8e957f | | WebShaper | 60.1 | Not recorded | 68fa7a941e209131cea84031 | | WebSailor | 55.4 | Not recorded | 68fa7a941e209131cea84031 | | WebShaper | 53.3 | Not recorded | 68fa7a941e209131cea84031 | | WebSailor | 53.2 | Not recorded | 68fa7a941e209131cea84031 | | WebDancer | 51.5 | Not recorded | 68fa7a941e209131cea84031 | | CSO | 49.5 | Not recorded | 6982d2aba8b01e76cc8e957f | | Cognitive Kernel-Pro | 49.3 | Not recorded | 68fa7a941e209131cea84031 | | WebThinker-RL | 48.5 | Not recorded | 68fa7a941e209131cea84031 | | GPT-4.1 | 45.6 | Not recorded | 6982d2aba8b01e76cc8e957f | | WebThinker-Base | 44.7 | Not recorded | 68fa7a941e209131cea84031 | | IPR | 44.6 | Not recorded | 6982d2aba8b01e76cc8e957f | | Cognitive Kernel-Pro | 43.7 | Not recorded | 68fa7a941e209131cea84031 | | Cognitive Kernel-Pro | 41.1 | Not recorded | 68fa7a941e209131cea84031 | | WebDancer | 40.7 | Not recorded | 68fa7a941e209131cea84031 | | ETO | 38.9 | Not recorded | 6982d2aba8b01e76cc8e957f | | Step-DPO | 38.9 | Not recorded | 6982d2aba8b01e76cc8e957f | | WebSailor | 37.9 | Not recorded | 68fa7a941e209131cea84031 | | CK-Pro-8B | 35.9 | Not recorded | 6982d2aba8b01e76cc8e957f | | RFT | 34.9 | Not recorded | 6982d2aba8b01e76cc8e957f | | WebDancer | 34 | Not recorded | 68fa7a941e209131cea84031 | | WebDancer | 31 | Not recorded | 68fa7a941e209131cea84031 | | Search-o1 | 28.2 | Not recorded | 68fa7a941e209131cea84031 | | Qwen3-8B | 20.4 | Not recorded | 6982d2aba8b01e76cc8e957f | | R1-Searcher | 20.4 | Not recorded | 68fa7a941e209131cea84031 | | Search-o1 | 17.5 | Not recorded | 68fa7a941e209131cea84031 | ### Claude-3.7-Sonnet Overall Accuracy: 62.1. Result ID: 69964fea5698665ec494aaa4. [Reporting paper]() Recorded evaluation configuration: {"Model Type":"Proprietary"}. This record does not assert independent reproduction. ### WebShaper Overall Accuracy: 60.1. Result ID: 69e9ebd547f34ce924d18cfa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"72B"}. This record does not assert independent reproduction. ### WebSailor Overall Accuracy: 55.4. Result ID: 69e9ebd547f34ce924d18cf8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"72B"}. This record does not assert independent reproduction. ### WebShaper Overall Accuracy: 53.3. Result ID: 69e9ebd547f34ce924d18cf9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"32B"}. This record does not assert independent reproduction. ### WebSailor Overall Accuracy: 53.2. Result ID: 69e9ebd547f34ce924d18cf7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"32B"}. This record does not assert independent reproduction. ### WebDancer Overall Accuracy: 51.5. Result ID: 69e9ebd547f34ce924d18cf6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"32B","Evaluation Protocol":"Pass@1 (Alternative Config)"}. This record does not assert independent reproduction. ### CSO Overall Accuracy: 49.5. Result ID: 69964fea5698665ec494aaab. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"CK-Pro-8B","Post-Training Method":"Critical Step Optimization"}. This record does not assert independent reproduction. ### Cognitive Kernel-Pro Overall Accuracy: 49.3. Result ID: 69e9ebd547f34ce924d18d02. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"8B","Evaluation Protocol":"Pass@3"}. This record does not assert independent reproduction. ### WebThinker-RL Overall Accuracy: 48.5. Result ID: 69e9ebd547f34ce924d18cf3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"32B"}. This record does not assert independent reproduction. ### GPT-4.1 Overall Accuracy: 45.6. Result ID: 69964fea5698665ec494aaa3. [Reporting paper]() Recorded evaluation configuration: {"Model Type":"Proprietary"}. This record does not assert independent reproduction. ### WebThinker-Base Overall Accuracy: 44.7. Result ID: 69e9ebd547f34ce924d18cf2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"32B"}. This record does not assert independent reproduction. ### IPR Overall Accuracy: 44.6. Result ID: 69964fea5698665ec494aaaa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"CK-Pro-8B","Post-Training Method":"Iterative Process Refinement"}. This record does not assert independent reproduction. ### Cognitive Kernel-Pro Overall Accuracy: 43.7. Result ID: 69e9ebd547f34ce924d18d00. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"8B","Backbone":"Qwen-3-8B","Evaluation Protocol":"Pass@1"}. This record does not assert independent reproduction. ### Cognitive Kernel-Pro Overall Accuracy: 41.1. Result ID: 69e9ebd547f34ce924d18d01. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"8B","Evaluation Protocol":"Voting"}. This record does not assert independent reproduction. ### WebDancer Overall Accuracy: 40.7. Result ID: 69e9ebd547f34ce924d18cf5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"32B","Evaluation Protocol":"Pass@1"}. This record does not assert independent reproduction. ### ETO Overall Accuracy: 38.9. Result ID: 69964fea5698665ec494aaa7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"CK-Pro-8B","Post-Training Method":"Exploration-based Trajectory Optimization"}. This record does not assert independent reproduction. ### Step-DPO Overall Accuracy: 38.9. Result ID: 69964fea5698665ec494aaa9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"CK-Pro-8B","Post-Training Method":"Step-wise DPO"}. This record does not assert independent reproduction. ### WebSailor Overall Accuracy: 37.9. Result ID: 69e9ebd547f34ce924d18cff. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"7B"}. This record does not assert independent reproduction. ### CK-Pro-8B Overall Accuracy: 35.9. Result ID: 69964fea5698665ec494aaa6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Mode":"SFT"}. This record does not assert independent reproduction. ### RFT Overall Accuracy: 34.9. Result ID: 69964fea5698665ec494aaa8. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Base Model":"CK-Pro-8B","Post-Training Method":"Rejection Sampling Fine-Tuning"}. This record does not assert independent reproduction. ### WebDancer Overall Accuracy: 34. Result ID: 69e9ebd547f34ce924d18cfe. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"7B","Evaluation Protocol":"Pass@3"}. This record does not assert independent reproduction. ### WebDancer Overall Accuracy: 31. Result ID: 69e9ebd547f34ce924d18cfd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"7B","Evaluation Protocol":"Pass@1"}. This record does not assert independent reproduction. ### Search-o1 Overall Accuracy: 28.2. Result ID: 69e9ebd547f34ce924d18cf4. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Size":"32B"}. This record does not assert independent reproduction. ### Qwen3-8B Overall Accuracy: 20.4. Result ID: 69964fea5698665ec494aaa5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Type":"Open-Source Baseline"}. This record does not assert independent reproduction. ### R1-Searcher Overall Accuracy: 20.4. Result ID: 69e9ebd547f34ce924d18cfc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"7B"}. This record does not assert independent reproduction. ### Search-o1 Overall Accuracy: 17.5. Result ID: 69e9ebd547f34ce924d18cfb. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Size":"7B"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.