# Multimodal Tool-Use on TIR-Bench Source: [Multimodal Tool-Use on TIR-Bench]() Collection index: [llms.txt]() Data updated: 2026-05-12T10:11:07.507Z Evaluation of multimodal model performance on tool utilization and reasoning capabilities using the TIR-Bench dataset, measured using Avg@4 and Pass@4 metrics. 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 | Multimodal Tool-Use | | Dataset (including any stated split) | TIR-Bench | | Metric | Avg@4 | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 16 of 16 | | 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 - [Avg@4](): Higher is better; unit: PERCENT. - [Pass@4](): Higher is better; unit: PERCENT. ## Recorded results | Method | Avg@4 | Result date | Report paper | | --- | --- | --- | --- | | Qwen3.5-35B-A3B | 50 | Not recorded | 6a02b408dbc836b87d824620 | | Qwen3.5-35B-A3B | 40.62 | Not recorded | 6a02b408dbc836b87d824620 | | Qwen3.5-Plus | 40.25 | Not recorded | 6a02b408dbc836b87d824620 | | Qwen3.5-35B-A3B | 39.37 | Not recorded | 6a02b408dbc836b87d824620 | | Qwen3.5-Plus | 29.88 | Not recorded | 6a02b408dbc836b87d824620 | | Gemini-2.5-Pro | 28.38 | Not recorded | 6a02b408dbc836b87d824620 | | o4-mini | 24.62 | Not recorded | 6a02b408dbc836b87d824620 | | GPT-5-mini | 23.5 | Not recorded | 6a02b408dbc836b87d824620 | | Qwen3.5-35B-A3B | 23 | Not recorded | 6a02b408dbc836b87d824620 | | Gemini-2.5-Pro | 21.37 | Not recorded | 6a02b408dbc836b87d824620 | | o4-mini | 20.13 | Not recorded | 6a02b408dbc836b87d824620 | | GPT-5-mini | 20 | Not recorded | 6a02b408dbc836b87d824620 | | Kimi-K2.5 | 19.38 | Not recorded | 6a02b408dbc836b87d824620 | | GLM-4.6V | 18.75 | Not recorded | 6a02b408dbc836b87d824620 | | GLM-4.6V | 17.5 | Not recorded | 6a02b408dbc836b87d824620 | | Kimi-K2.5 | 17.37 | Not recorded | 6a02b408dbc836b87d824620 | ### Qwen3.5-35B-A3B Avg@4: 50. Result ID: 6a02fc1347f34ce924e64b8a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"w/ ACE-SKILL"}. This record does not assert independent reproduction. ### Qwen3.5-35B-A3B Avg@4: 40.62. Result ID: 6a02fc1347f34ce924e64b89. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"w/ XSkill"}. This record does not assert independent reproduction. ### Qwen3.5-Plus Avg@4: 40.25. Result ID: 6a02fc1347f34ce924e64b82. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Vanilla Tools"}. This record does not assert independent reproduction. ### Qwen3.5-35B-A3B Avg@4: 39.37. Result ID: 6a02fc1347f34ce924e64b88. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Vanilla Tools"}. This record does not assert independent reproduction. ### Qwen3.5-Plus Avg@4: 29.88. Result ID: 6a02fc1347f34ce924e64b81. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Native Direct Answer"}. This record does not assert independent reproduction. ### Gemini-2.5-Pro Avg@4: 28.38. Result ID: 6a02fc1347f34ce924e64b80. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Vanilla Tools"}. This record does not assert independent reproduction. ### o4-mini Avg@4: 24.62. Result ID: 6a02fc1347f34ce924e64b7c. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Vanilla Tools"}. This record does not assert independent reproduction. ### GPT-5-mini Avg@4: 23.5. Result ID: 6a02fc1347f34ce924e64b7e. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Vanilla Tools"}. This record does not assert independent reproduction. ### Qwen3.5-35B-A3B Avg@4: 23. Result ID: 6a02fc1347f34ce924e64b87. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Native Direct Answer"}. This record does not assert independent reproduction. ### Gemini-2.5-Pro Avg@4: 21.37. Result ID: 6a02fc1347f34ce924e64b7f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Native Direct Answer"}. This record does not assert independent reproduction. ### o4-mini Avg@4: 20.13. Result ID: 6a02fc1347f34ce924e64b7b. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Native Direct Answer"}. This record does not assert independent reproduction. ### GPT-5-mini Avg@4: 20. Result ID: 6a02fc1347f34ce924e64b7d. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Native Direct Answer"}. This record does not assert independent reproduction. ### Kimi-K2.5 Avg@4: 19.38. Result ID: 6a02fc1347f34ce924e64b86. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Evaluation Protocol":"Vanilla Tools"}. This record does not assert independent reproduction. ### GLM-4.6V Avg@4: 18.75. Result ID: 6a02fc1347f34ce924e64b84. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Vanilla Tools"}. This record does not assert independent reproduction. ### GLM-4.6V Avg@4: 17.5. Result ID: 6a02fc1347f34ce924e64b83. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Native Direct Answer"}. This record does not assert independent reproduction. ### Kimi-K2.5 Avg@4: 17.37. Result ID: 6a02fc1347f34ce924e64b85. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Evaluation Protocol":"Native Direct Answer"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.