# Multi-image Grounding on MIG-Bench 1.0 (full) Source: [Multi-image Grounding on MIG-Bench 1.0 (full)]() Collection index: [llms.txt]() Data updated: 2026-04-27T07:23:32.445Z Evaluation of Multi-image Grounding performance on the full split of the MIG-Bench 1.0 dataset, assessed using specialized metrics including Static, Robust, Common, OT, MV, Region, Refer, GG, Reason, Co-Re, and AVE scores. 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 | Multi-image Grounding | | Dataset (including any stated split) | MIG-Bench 1.0 (full) | | Metric | Static Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 19 of 19 | | 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 - [Static Score](): Higher is better; unit: PERCENT. - [Robust Score](): Higher is better; unit: PERCENT. - [Common Score](): Higher is better; unit: PERCENT. - [OT Score](): Higher is better; unit: PERCENT. - [MV Score](): Higher is better; unit: PERCENT. - [Region Score](): Higher is better; unit: PERCENT. - [Refer Score](): Higher is better; unit: PERCENT. - [GG Score](): Higher is better; unit: PERCENT. - [Reason Score](): Higher is better; unit: PERCENT. - [Co-Re Score](): Higher is better; unit: PERCENT. - [AVE Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Static Score | Result date | Report paper | | --- | --- | --- | --- | | Human | 99.5 | Not recorded | 69eeed6355a3c32af89f95fa | | Migician-7B | 65.15 | Not recorded | 69eeed6355a3c32af89f95fa | | CGC-8B | 55.3 | Not recorded | 69eeed6355a3c32af89f95fa | | CGC-7B | 48.86 | Not recorded | 69eeed6355a3c32af89f95fa | | Qwen2-VL-72B | 46.12 | Not recorded | 69eeed6355a3c32af89f95fa | | Qwen2.5-VL-72B | 43.75 | Not recorded | 69eeed6355a3c32af89f95fa | | Qwen3-VL-8B | 41.1 | Not recorded | 69eeed6355a3c32af89f95fa | | Visual-RFT-7B | 33.95 | Not recorded | 69eeed6355a3c32af89f95fa | | Qwen2-VL-7B | 27.84 | Not recorded | 69eeed6355a3c32af89f95fa | | InternVL3-8B | 23.67 | Not recorded | 69eeed6355a3c32af89f95fa | | mPLUG-Owl3-8B | 18.56 | Not recorded | 69eeed6355a3c32af89f95fa | | InternVL2-76B | 15.91 | Not recorded | 69eeed6355a3c32af89f95fa | | MiniCPM2.6-8B | 14.58 | Not recorded | 69eeed6355a3c32af89f95fa | | LLaVA-OV-72B | 13.26 | Not recorded | 69eeed6355a3c32af89f95fa | | InternVL3-78B | 10.04 | Not recorded | 69eeed6355a3c32af89f95fa | | Qwen2.5-VL-7B | 9.57 | Not recorded | 69eeed6355a3c32af89f95fa | | InternVL2-8B | 6.92 | Not recorded | 69eeed6355a3c32af89f95fa | | LLaVA-OV-7B | 6.06 | Not recorded | 69eeed6355a3c32af89f95fa | | Mantis-8B | 1.52 | Not recorded | 69eeed6355a3c32af89f95fa | ### Human Static Score: 99.5. Result ID: 69ef0edd47f34ce924d4dec1. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Migician-7B Static Score: 65.15. Result ID: 69ef0edd47f34ce924d4ded1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base MLLM":"Qwen2-VL"}. This record does not assert independent reproduction. ### CGC-8B Static Score: 55.3. Result ID: 69ef0edd47f34ce924d4ded3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base MLLM":"Qwen3-VL"}. This record does not assert independent reproduction. ### CGC-7B Static Score: 48.86. Result ID: 69ef0edd47f34ce924d4ded2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base MLLM":"Qwen2-VL"}. This record does not assert independent reproduction. ### Qwen2-VL-72B Static Score: 46.12. Result ID: 69ef0edd47f34ce924d4dec5. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"Large-Scale Models (≥70B)"}. This record does not assert independent reproduction. ### Qwen2.5-VL-72B Static Score: 43.75. Result ID: 69ef0edd47f34ce924d4dec6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"Large-Scale Models (≥70B)"}. This record does not assert independent reproduction. ### Qwen3-VL-8B Static Score: 41.1. Result ID: 69ef0edd47f34ce924d4decf. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"7 ∼ 8B Models"}. This record does not assert independent reproduction. ### Visual-RFT-7B Static Score: 33.95. Result ID: 69ef0edd47f34ce924d4ded0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base MLLM":"Qwen2-VL"}. This record does not assert independent reproduction. ### Qwen2-VL-7B Static Score: 27.84. Result ID: 69ef0edd47f34ce924d4decd. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"7 ∼ 8B Models"}. This record does not assert independent reproduction. ### InternVL3-8B Static Score: 23.67. Result ID: 69ef0edd47f34ce924d4decc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"7 ∼ 8B Models"}. This record does not assert independent reproduction. ### mPLUG-Owl3-8B Static Score: 18.56. Result ID: 69ef0edd47f34ce924d4deca. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"7 ∼ 8B Models"}. This record does not assert independent reproduction. ### InternVL2-76B Static Score: 15.91. Result ID: 69ef0edd47f34ce924d4dec3. [Reporting paper]() Recorded evaluation configuration: {"Model Scale":"Large-Scale Models (≥70B)"}. This record does not assert independent reproduction. ### MiniCPM2.6-8B Static Score: 14.58. Result ID: 69ef0edd47f34ce924d4dec9. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"7 ∼ 8B Models"}. This record does not assert independent reproduction. ### LLaVA-OV-72B Static Score: 13.26. Result ID: 69ef0edd47f34ce924d4dec2. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"Large-Scale Models (≥70B)"}. This record does not assert independent reproduction. ### InternVL3-78B Static Score: 10.04. Result ID: 69ef0edd47f34ce924d4dec4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"Large-Scale Models (≥70B)"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B Static Score: 9.57. Result ID: 69ef0edd47f34ce924d4dece. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"7 ∼ 8B Models"}. This record does not assert independent reproduction. ### InternVL2-8B Static Score: 6.92. Result ID: 69ef0edd47f34ce924d4decb. [Reporting paper]() Recorded evaluation configuration: {"Model Scale":"7 ∼ 8B Models"}. This record does not assert independent reproduction. ### LLaVA-OV-7B Static Score: 6.06. Result ID: 69ef0edd47f34ce924d4dec8. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"7 ∼ 8B Models"}. This record does not assert independent reproduction. ### Mantis-8B Static Score: 1.52. Result ID: 69ef0edd47f34ce924d4dec7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"7 ∼ 8B Models"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.