# Vision-Language Perception on VL Perception Benchmarks TextVQA, POPE, Seed-Bench

Source: [Vision-Language Perception on VL Perception Benchmarks TextVQA, POPE, Seed-Bench](<https://www.sota2.com/research/sota/vision-language-perception-on-vl-perception-benchmarks-textvqa-pope-seed-bench>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T06:06:22.816Z

Evaluation of model performance across the VL Perception suite, including TextVQA, POPE, and Seed-Bench tasks.

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 | Vision-Language Perception |
| Dataset (including any stated split) | VL Perception Benchmarks TextVQA, POPE, Seed-Bench |
| Metric | TextVQA Score |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 25 of 25 |
| 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

- [TextVQA Score](<https://www.sota2.com/research/sota/vision-language-perception-on-vl-perception-benchmarks-textvqa-pope-seed-bench.md?metric=TextVQA%20Score>): Higher is better; unit: PERCENT.

- [POPE Score](<https://www.sota2.com/research/sota/vision-language-perception-on-vl-perception-benchmarks-textvqa-pope-seed-bench.md?metric=POPE%20Score>): Higher is better; unit: PERCENT.

- [Seed-Bench Score](<https://www.sota2.com/research/sota/vision-language-perception-on-vl-perception-benchmarks-textvqa-pope-seed-bench.md?metric=Seed-Bench%20Score>): Higher is better; unit: PERCENT.

- [Average Score](<https://www.sota2.com/research/sota/vision-language-perception-on-vl-perception-benchmarks-textvqa-pope-seed-bench.md?metric=Average%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | TextVQA Score | Result date | Report paper |
| --- | --- | --- | --- |
| Base | 85.5 | Not recorded | 697c3aefa8b01e76cc8da38e |
| FRISM | 85 | Not recorded | 697c3aefa8b01e76cc8da38e |
| DARE | 84.9 | Not recorded | 697c3aefa8b01e76cc8da38e |
| TA | 83.9 | Not recorded | 697c3aefa8b01e76cc8da38e |
| IP-Merging | 82.9 | Not recorded | 697c3aefa8b01e76cc8da38e |
| IP-Merging | 82 | Not recorded | 697c3aefa8b01e76cc8da38e |
| IP-Merging | 81.6 | Not recorded | 697c3aefa8b01e76cc8da38e |
| TA | 79.9 | Not recorded | 697c3aefa8b01e76cc8da38e |
| FRISM | 79.9 | Not recorded | 697c3aefa8b01e76cc8da38e |
| Base | 79.8 | Not recorded | 697c3aefa8b01e76cc8da38e |
| Base | 79.3 | Not recorded | 697c3aefa8b01e76cc8da38e |
| TA | 79.2 | Not recorded | 697c3aefa8b01e76cc8da38e |
| DARE | 79.2 | Not recorded | 697c3aefa8b01e76cc8da38e |
| FRISM | 79.2 | Not recorded | 697c3aefa8b01e76cc8da38e |
| TA | 79.1 | Not recorded | 697c3aefa8b01e76cc8da38e |
| TA | 79 | Not recorded | 697c3aefa8b01e76cc8da38e |
| Ties-Merging | 78.3 | Not recorded | 697c3aefa8b01e76cc8da38e |
| Ties-Merging | 77.9 | Not recorded | 697c3aefa8b01e76cc8da38e |
| IP-Merging | 76.9 | Not recorded | 697c3aefa8b01e76cc8da38e |
| IP-Merging | 76.9 | Not recorded | 697c3aefa8b01e76cc8da38e |
| Ties-Merging | 76.4 | Not recorded | 697c3aefa8b01e76cc8da38e |
| IP-Merging | 75.8 | Not recorded | 697c3aefa8b01e76cc8da38e |
| IP-Merging | 70.6 | Not recorded | 697c3aefa8b01e76cc8da38e |
| TA | 69.1 | Not recorded | 697c3aefa8b01e76cc8da38e |
| TA | 68 | Not recorded | 697c3aefa8b01e76cc8da38e |

### Base

TextVQA Score: 85.5. Result ID: 69977dfe5698665ec4d28665.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-7B-Instruct","Reasoning Model":"DeepSeek-R1-Distill-Qwen-7B"}. This record does not assert independent reproduction.

### FRISM

TextVQA Score: 85. Result ID: 69977dfe5698665ec4d2866e.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-7B-Instruct","Reasoning Model":"DeepSeek-R1-Distill-Qwen-7B"}. This record does not assert independent reproduction.

### DARE

TextVQA Score: 84.9. Result ID: 69977dfe5698665ec4d2866a.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73921e209131cea5d4aa.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-7B-Instruct","Reasoning Model":"DeepSeek-R1-Distill-Qwen-7B"}. This record does not assert independent reproduction.

### TA

TextVQA Score: 83.9. Result ID: 69977dfe5698665ec4d28666.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ef01e209131cea43012.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-7B-Instruct","Reasoning Model":"DeepSeek-R1-Distill-Qwen-7B","Lambda (λ)":"0.1"}. This record does not assert independent reproduction.

### IP-Merging

TextVQA Score: 82.9. Result ID: 69977dfe5698665ec4d2866d.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c281e209131cea8cbb8.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-7B-Instruct","Reasoning Model":"DeepSeek-R1-Distill-Qwen-7B","Temperature (T)":"0.4"}. This record does not assert independent reproduction.

### IP-Merging

TextVQA Score: 82. Result ID: 69977dfe5698665ec4d2866c.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c281e209131cea8cbb8.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-7B-Instruct","Reasoning Model":"DeepSeek-R1-Distill-Qwen-7B","Temperature (T)":"0.3"}. This record does not assert independent reproduction.

### IP-Merging

TextVQA Score: 81.6. Result ID: 69977dfe5698665ec4d2866b.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c281e209131cea8cbb8.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-7B-Instruct","Reasoning Model":"DeepSeek-R1-Distill-Qwen-7B","Temperature (T)":"0.2"}. This record does not assert independent reproduction.

### TA

TextVQA Score: 79.9. Result ID: 69977dfe5698665ec4d28670.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ef01e209131cea43012.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-32B-Instruct","Reasoning Model":"QwQ-32B"}. This record does not assert independent reproduction.

### FRISM

TextVQA Score: 79.9. Result ID: 69977dfe5698665ec4d28673.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-32B-Instruct","Reasoning Model":"QwQ-32B"}. This record does not assert independent reproduction.

### Base

TextVQA Score: 79.8. Result ID: 69977dfe5698665ec4d2866f.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-32B-Instruct","Reasoning Model":"QwQ-32B"}. This record does not assert independent reproduction.

### Base

TextVQA Score: 79.3. Result ID: 69977dfe5698665ec4d2865b.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-3B-Instruct","Reasoning Model":"SmallThinker-3B"}. This record does not assert independent reproduction.

### TA

TextVQA Score: 79.2. Result ID: 69977dfe5698665ec4d2865c.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ef01e209131cea43012.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-3B-Instruct","Reasoning Model":"SmallThinker-3B","Lambda (λ)":"0.1"}. This record does not assert independent reproduction.

### DARE

TextVQA Score: 79.2. Result ID: 69977dfe5698665ec4d28660.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73921e209131cea5d4aa.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-3B-Instruct","Reasoning Model":"SmallThinker-3B"}. This record does not assert independent reproduction.

### FRISM

TextVQA Score: 79.2. Result ID: 69977dfe5698665ec4d28664.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-3B-Instruct","Reasoning Model":"SmallThinker-3B"}. This record does not assert independent reproduction.

### TA

TextVQA Score: 79.1. Result ID: 69977dfe5698665ec4d2865d.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ef01e209131cea43012.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-3B-Instruct","Reasoning Model":"SmallThinker-3B","Lambda (λ)":"0.15"}. This record does not assert independent reproduction.

### TA

TextVQA Score: 79. Result ID: 69977dfe5698665ec4d2865e.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ef01e209131cea43012.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-3B-Instruct","Reasoning Model":"SmallThinker-3B","Lambda (λ)":"0.2"}. This record does not assert independent reproduction.

### Ties-Merging

TextVQA Score: 78.3. Result ID: 69977dfe5698665ec4d28671.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71091e209131cea4ecd8.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-32B-Instruct","Reasoning Model":"QwQ-32B"}. This record does not assert independent reproduction.

### Ties-Merging

TextVQA Score: 77.9. Result ID: 69977dfe5698665ec4d28669.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71091e209131cea4ecd8.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-7B-Instruct","Reasoning Model":"DeepSeek-R1-Distill-Qwen-7B"}. This record does not assert independent reproduction.

### IP-Merging

TextVQA Score: 76.9. Result ID: 69977dfe5698665ec4d28662.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c281e209131cea8cbb8.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-3B-Instruct","Reasoning Model":"SmallThinker-3B","Temperature (T)":"0.3"}. This record does not assert independent reproduction.

### IP-Merging

TextVQA Score: 76.9. Result ID: 69977dfe5698665ec4d28663.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c281e209131cea8cbb8.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-3B-Instruct","Reasoning Model":"SmallThinker-3B","Temperature (T)":"0.4"}. This record does not assert independent reproduction.

### Ties-Merging

TextVQA Score: 76.4. Result ID: 69977dfe5698665ec4d2865f.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71091e209131cea4ecd8.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-3B-Instruct","Reasoning Model":"SmallThinker-3B"}. This record does not assert independent reproduction.

### IP-Merging

TextVQA Score: 75.8. Result ID: 69977dfe5698665ec4d28672.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c281e209131cea8cbb8.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-32B-Instruct","Reasoning Model":"QwQ-32B"}. This record does not assert independent reproduction.

### IP-Merging

TextVQA Score: 70.6. Result ID: 69977dfe5698665ec4d28661.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c281e209131cea8cbb8.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-3B-Instruct","Reasoning Model":"SmallThinker-3B","Temperature (T)":"0.2"}. This record does not assert independent reproduction.

### TA

TextVQA Score: 69.1. Result ID: 69977dfe5698665ec4d28667.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ef01e209131cea43012.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-7B-Instruct","Reasoning Model":"DeepSeek-R1-Distill-Qwen-7B","Lambda (λ)":"0.15"}. This record does not assert independent reproduction.

### TA

TextVQA Score: 68. Result ID: 69977dfe5698665ec4d28668.

[Reporting paper](<https://www.sota2.com/research/paper/697c3aefa8b01e76cc8da38e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ef01e209131cea43012.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-VL-7B-Instruct","Reasoning Model":"DeepSeek-R1-Distill-Qwen-7B","Lambda (λ)":"0.2"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
