# Mathematical Reasoning on GAOKAO-MM Math Source: [Mathematical Reasoning on GAOKAO-MM Math]() Collection index: [llms.txt]() Data updated: 2026-03-25T13:06:23.852Z Evaluation of model performance on mathematical reasoning tasks using the GAOKAO-MM Math dataset, measured by Accuracy. 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 | Mathematical Reasoning | | Dataset (including any stated split) | GAOKAO-MM Math | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 17 of 17 | | 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 - [Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | DS-R1-Distill-Qwen-32B + GeoTikzBridge-Base | 68.8 | Not recorded | 69c36c000636e5bf4afc6464 | | GLM-4.5-Air-106B-A12B + GeoTikzBridge-Base | 68.8 | Not recorded | 69c36c000636e5bf4afc6464 | | GLM-Z1-32B + GeoTikzBridge-Base | 67.5 | Not recorded | 69c36c000636e5bf4afc6464 | | GLM4.5-V-106B-A12B + GeoTikzBridge-Base | 66.3 | Not recorded | 69c36c000636e5bf4afc6464 | | Skywork-OR1-32B + GeoTikzBridge-Base | 66.3 | Not recorded | 69c36c000636e5bf4afc6464 | | InternVL3-78B | 63.8 | Not recorded | 69c36c000636e5bf4afc6464 | | InternVL3.5-38B + GeoTikzBridge-Base | 63.5 | Not recorded | 69c36c000636e5bf4afc6464 | | GPT-OSS-120B + GeoTikzBridge-Base | 62.5 | Not recorded | 69c36c000636e5bf4afc6464 | | Qwen2.5-VL-72B | 61.3 | Not recorded | 69c36c000636e5bf4afc6464 | | GLM4.5-V-106B-A12B | 61.3 | Not recorded | 69c36c000636e5bf4afc6464 | | InternVL3.5-38B | 60 | Not recorded | 69c36c000636e5bf4afc6464 | | Qwen3-VL-30B-A3B + GeoTikzBridge-Base | 58.8 | Not recorded | 69c36c000636e5bf4afc6464 | | Qwen3-VL-30B-A3B | 55 | Not recorded | 69c36c000636e5bf4afc6464 | | Qwen3-30B-A3B + GeoTikzBridge-Base | 55 | Not recorded | 69c36c000636e5bf4afc6464 | | GPT-4-turbo | 50 | Not recorded | 69c36c000636e5bf4afc6464 | | Qwen2.5-VL-32B | 50 | Not recorded | 69c36c000636e5bf4afc6464 | | GPT-4V | 45 | Not recorded | 69c36c000636e5bf4afc6464 | ### DS-R1-Distill-Qwen-32B + GeoTikzBridge-Base Accuracy: 68.8. Result ID: 69c3dd8147f34ce924ad4145. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tikz-supplementation":"true"}. This record does not assert independent reproduction. ### GLM-4.5-Air-106B-A12B + GeoTikzBridge-Base Accuracy: 68.8. Result ID: 69c3dd8147f34ce924ad4146. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tikz-supplementation":"true"}. This record does not assert independent reproduction. ### GLM-Z1-32B + GeoTikzBridge-Base Accuracy: 67.5. Result ID: 69c3dd8147f34ce924ad4144. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tikz-supplementation":"true"}. This record does not assert independent reproduction. ### GLM4.5-V-106B-A12B + GeoTikzBridge-Base Accuracy: 66.3. Result ID: 69c3dd8147f34ce924ad4141. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tikz-supplementation":"true"}. This record does not assert independent reproduction. ### Skywork-OR1-32B + GeoTikzBridge-Base Accuracy: 66.3. Result ID: 69c3dd8147f34ce924ad4142. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tikz-supplementation":"true"}. This record does not assert independent reproduction. ### InternVL3-78B Accuracy: 63.8. Result ID: 69c3dd8147f34ce924ad413b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tikz-supplementation":"false"}. This record does not assert independent reproduction. ### InternVL3.5-38B + GeoTikzBridge-Base Accuracy: 63.5. Result ID: 69c3dd8147f34ce924ad413f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tikz-supplementation":"true"}. This record does not assert independent reproduction. ### GPT-OSS-120B + GeoTikzBridge-Base Accuracy: 62.5. Result ID: 69c3dd8147f34ce924ad4147. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tikz-supplementation":"true"}. This record does not assert independent reproduction. ### Qwen2.5-VL-72B Accuracy: 61.3. Result ID: 69c3dd8147f34ce924ad413a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tikz-supplementation":"false"}. This record does not assert independent reproduction. ### GLM4.5-V-106B-A12B Accuracy: 61.3. Result ID: 69c3dd8147f34ce924ad4140. [Reporting paper]() Recorded evaluation configuration: {"Tikz-supplementation":"false"}. This record does not assert independent reproduction. ### InternVL3.5-38B Accuracy: 60. Result ID: 69c3dd8147f34ce924ad413e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tikz-supplementation":"false"}. This record does not assert independent reproduction. ### Qwen3-VL-30B-A3B + GeoTikzBridge-Base Accuracy: 58.8. Result ID: 69c3dd8147f34ce924ad413d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tikz-supplementation":"true"}. This record does not assert independent reproduction. ### Qwen3-VL-30B-A3B Accuracy: 55. Result ID: 69c3dd8147f34ce924ad413c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tikz-supplementation":"false"}. This record does not assert independent reproduction. ### Qwen3-30B-A3B + GeoTikzBridge-Base Accuracy: 55. Result ID: 69c3dd8147f34ce924ad4143. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tikz-supplementation":"true"}. This record does not assert independent reproduction. ### GPT-4-turbo Accuracy: 50. Result ID: 69c3dd8147f34ce924ad4138. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Tikz-supplementation":"false"}. This record does not assert independent reproduction. ### Qwen2.5-VL-32B Accuracy: 50. Result ID: 69c3dd8147f34ce924ad4139. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tikz-supplementation":"false"}. This record does not assert independent reproduction. ### GPT-4V Accuracy: 45. Result ID: 69c3dd8147f34ce924ad4137. [Reporting paper]() Recorded evaluation configuration: {"Tikz-supplementation":"false"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.