# High-Resolution Visual Reasoning on HR-Bench (4K/8K configurations) Source: [High-Resolution Visual Reasoning on HR-Bench (4K/8K configurations)]() Collection index: [llms.txt]() Data updated: 2026-06-01T07:33:47.618Z Evaluation of model performance on the HR-Bench dataset for High-Resolution Visual Reasoning, specifically reporting results across 4K and 8K resolution configurations, along with the average score. 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 | High-Resolution Visual Reasoning | | Dataset (including any stated split) | HR-Bench | | Metric | Score (4K) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 30 of 30 | | 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 - [Score (4K)](): Higher is better; unit: PERCENT. - [Score (8K)](): Higher is better; unit: PERCENT. - [Average Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Score (4K) | Result date | Report paper | | --- | --- | --- | --- | | MAPO | 81 | Not recorded | 69d732b77d0172140ec76cda | | Ovis2.5-9B + GRPO | 78.5 | Not recorded | 69d732b77d0172140ec76cda | | Ovis2.5-9B + GSPO | 78.3 | Not recorded | 69d732b77d0172140ec76cda | | Ovis2.5-9B + DAPO | 78.1 | Not recorded | 69d732b77d0172140ec76cda | | DeepEyesv2 | 77.9 | Not recorded | 69d732b77d0172140ec76cda | | Ovis2.5-9B + PPO | 77.4 | Not recorded | 69d732b77d0172140ec76cda | | Thyme | 77 | Not recorded | 69d732b77d0172140ec76cda | | Mini-o3 | 75.4 | Not recorded | 69d732b77d0172140ec76cda | | Deepeyes | 75 | Not recorded | 693a50176cfd488761a54aa5 | | DeepEyes | 74.9 | Not recorded | 69d732b77d0172140ec76cda | | GPT-5 | 74.6 | Not recorded | 69d732b77d0172140ec76cda | | Ovis2.5-9B + Coldstart SFT | 74.6 | Not recorded | 69d732b77d0172140ec76cda | | Ovis2.5-9B | 74 | Not recorded | 69d732b77d0172140ec76cda | | IMAgent | 73.5 | Not recorded | 693a50176cfd488761a54aa5 | | Qwen2.5-VL | 70.62 | Not recorded | 693a50176cfd488761a54aa5 | | ReGuLaR-7B | 70.5 | Not recorded | 6a1d120bdbd7150506f83f6d | | Laser | 70.25 | Not recorded | 6a1d120bdbd7150506f83f6d | | Zoomeye | 69.6 | Not recorded | 693a50176cfd488761a54aa5 | | Deepeyes | 69.25 | Not recorded | 6a1d120bdbd7150506f83f6d | | Pixel Reasoner | 68.75 | Not recorded | 693a50176cfd488761a54aa5 | | Qwen2.5-VL-CoT | 68.5 | Not recorded | 693a50176cfd488761a54aa5 | | Qwen2.5-VL-7B | 68 | Not recorded | 6a1d120bdbd7150506f83f6d | | Monet | 67.37 | Not recorded | 6a1d120bdbd7150506f83f6d | | Gemini 2.5 Pro | 67.3 | Not recorded | 69d732b77d0172140ec76cda | | Vision-R1 | 66.63 | Not recorded | 6a1d120bdbd7150506f83f6d | | LLaVA-OneVision | 63.88 | Not recorded | 6a1d120bdbd7150506f83f6d | | LLaVA-OneVision | 63 | Not recorded | 693a50176cfd488761a54aa5 | | GPT-4o | 59 | Not recorded | 693a50176cfd488761a54aa5 | | GPT-4o | 59 | Not recorded | 6a1d120bdbd7150506f83f6d | | LVR | 51.13 | Not recorded | 6a1d120bdbd7150506f83f6d | ### MAPO Score (4K): 81. Result ID: 69d8421447f34ce924c04b64. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"backbone":"Ovis2.5-9B"}. This record does not assert independent reproduction. ### Ovis2.5-9B + GRPO Score (4K): 78.5. Result ID: 69d8421447f34ce924c04b61. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"training":"GRPO"}. This record does not assert independent reproduction. ### Ovis2.5-9B + GSPO Score (4K): 78.3. Result ID: 69d8421447f34ce924c04b63. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"training":"GSPO"}. This record does not assert independent reproduction. ### Ovis2.5-9B + DAPO Score (4K): 78.1. Result ID: 69d8421447f34ce924c04b62. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"training":"DAPO"}. This record does not assert independent reproduction. ### DeepEyesv2 Score (4K): 77.9. Result ID: 69d8421447f34ce924c04b5d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"results\_source":"original paper"}. This record does not assert independent reproduction. ### Ovis2.5-9B + PPO Score (4K): 77.4. Result ID: 69d8421447f34ce924c04b60. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"training":"PPO"}. This record does not assert independent reproduction. ### Thyme Score (4K): 77. Result ID: 69d8421447f34ce924c04b5b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"results\_source":"original paper"}. This record does not assert independent reproduction. ### Mini-o3 Score (4K): 75.4. Result ID: 69d8421447f34ce924c04b5c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"reproduced\_by\_authors":"true"}. This record does not assert independent reproduction. ### Deepeyes Score (4K): 75. Result ID: 699889005698665ec4ea7fff. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"E2E":true,"Param":"7B"}. This record does not assert independent reproduction. ### DeepEyes Score (4K): 74.9. Result ID: 69d8421447f34ce924c04b5a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"reproduced\_by\_authors":"true"}. This record does not assert independent reproduction. ### GPT-5 Score (4K): 74.6. Result ID: 69d8421447f34ce924c04b58. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Ovis2.5-9B + Coldstart SFT Score (4K): 74.6. Result ID: 69d8421447f34ce924c04b5f. [Reporting paper]() Recorded evaluation configuration: {"training":"Supervised Fine-Tuning"}. This record does not assert independent reproduction. ### Ovis2.5-9B Score (4K): 74. Result ID: 69d8421447f34ce924c04b5e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### IMAgent Score (4K): 73.5. Result ID: 699889005698665ec4ea8001. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"E2E":true,"Param":"7B"}. This record does not assert independent reproduction. ### Qwen2.5-VL Score (4K): 70.62. Result ID: 699889005698665ec4ea7ffd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"E2E":null,"Param":"7B"}. This record does not assert independent reproduction. ### ReGuLaR-7B Score (4K): 70.5. Result ID: 6a1d35cbfbc373592c160782. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"Latent reasoning LVLMs"}. This record does not assert independent reproduction. ### Laser Score (4K): 70.25. Result ID: 6a1d35cbfbc373592c160781. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"Latent reasoning LVLMs"}. This record does not assert independent reproduction. ### Zoomeye Score (4K): 69.6. Result ID: 699889005698665ec4ea7ffb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"E2E":false,"Param":"7B"}. This record does not assert independent reproduction. ### Deepeyes Score (4K): 69.25. Result ID: 6a1d35cbfbc373592c16077d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"RL-based reasoning LVLMs"}. This record does not assert independent reproduction. ### Pixel Reasoner Score (4K): 68.75. Result ID: 699889005698665ec4ea8000. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"E2E":false,"Param":"7B"}. This record does not assert independent reproduction. ### Qwen2.5-VL-CoT Score (4K): 68.5. Result ID: 699889005698665ec4ea7ffe. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"E2E":null,"Param":"7B"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B Score (4K): 68. Result ID: 6a1d35cbfbc373592c16077c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"General-purpose LVLMs"}. This record does not assert independent reproduction. ### Monet Score (4K): 67.37. Result ID: 6a1d35cbfbc373592c160780. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Category":"Latent reasoning LVLMs"}. This record does not assert independent reproduction. ### Gemini 2.5 Pro Score (4K): 67.3. Result ID: 69d8421447f34ce924c04b59. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Vision-R1 Score (4K): 66.63. Result ID: 6a1d35cbfbc373592c16077e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"RL-based reasoning LVLMs"}. This record does not assert independent reproduction. ### LLaVA-OneVision Score (4K): 63.88. Result ID: 6a1d35cbfbc373592c16077b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Category":"General-purpose LVLMs"}. This record does not assert independent reproduction. ### LLaVA-OneVision Score (4K): 63. Result ID: 699889005698665ec4ea7ffc. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"E2E":null,"Param":"7B"}. This record does not assert independent reproduction. ### GPT-4o Score (4K): 59. Result ID: 699889005698665ec4ea7ffa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"E2E":null,"Param":null}. This record does not assert independent reproduction. ### GPT-4o Score (4K): 59. Result ID: 6a1d35cbfbc373592c16077a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Category":"General-purpose LVLMs"}. This record does not assert independent reproduction. ### LVR Score (4K): 51.13. Result ID: 6a1d35cbfbc373592c16077f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"Latent reasoning LVLMs"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.