# VisualToolBench (VTB)

> Source: Benchize (https://benchize.com) — https://benchize.com/benchmarks/scale%3Avtb. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/benchmarks/scale%3Avtb.

VisualToolBench (VTB)is the first benchmark designed to evaluate how well Multimodal Large Language Models (MLLMs) can dynamically interact with and reason about visual information. It shifts the paradigm from passively “thinkingaboutimages” to actively “thinkingwithimages,” treating them as a manipulable cognitive workspace. To solve complex, multi-step problems, models must use tools to transform visual content by cropping, editing, or enhancing it to uncover critical details. The benchmark provides leaderboard results across 16 diverse MLLMs, including reasoning, non-reasoning, open-source, and closed-source models. The VTB dataset was constructed to mirror the complexity of real-world user needs through the following features:

Canonical URL: https://benchize.com/benchmarks/scale%3Avtb
Last updated: Wed, 21 Oct 2026 18:52:20 GMT

Reproduction and quotation must name Benchize, link to the canonical HTML URL, and retain individual source citations.

VisualToolBench (VTB)is the first benchmark designed to evaluate how well Multimodal Large Language Models (MLLMs) can dynamically interact with and reason about visual information. It shifts the paradigm from passively “thinkingaboutimages” to actively “thinkingwithimages,” treating them as a manipulable cognitive workspace. To solve complex, multi-step problems, models must use tools to transform visual content by cropping, editing, or enhancing it to uncover critical details. The benchmark provides leaderboard results across 16 diverse MLLMs, including reasoning, non-reasoning, open-source, and closed-source models. The VTB dataset was constructed to mirror the complexity of real-world user needs through the following features:

Multimodal · Models tested: 5 · Score publishers: [scale](https://labs.scale.com/leaderboard/vtb)

## Model results

| # | Model | Score | ± |  |
| --- | --- | --- | --- | --- |
| 1 | [Muse Spark 1.1](https://benchize.com/models/muse-spark-1-1.md) | 44.77source score | CI 41.95–47.59 |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 44.77 \| source_score \| reported \| Higher is better \| ### Configuration Reasoning effort unknown Model configuration Muse Spark 1.1 sourceModelVersion Not reported thinkingEnabled Not reported ### Reported statistics Uncertainty Confidence interval: 41.95–47.59 contaminationWarning Not reported rank 1 reportedConfidenceHalfWidth 2.82 ### Source [Original source](https://labs.scale.com/leaderboard/vtb) Source scale Original model label Muse Spark 1.1 Observed 2026-10-21T16:35:36.362Z Evidence SHA-256 `70c39c91bb42d31fb0581b426c57134d54c74cc00ac658f5bd00694f52418c45` Source locator `/rows/0` |  |  |  |  |
| 2 | [Gemini 3.1 Pro Preview](https://benchize.com/models/gemini-3-1-pro-preview.md) | 28.97source score | CI 28.06–29.88 |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 28.97 \| source_score \| reported \| Higher is better \| ### Configuration Reasoning effort unknown Model configuration gemini-3.1-pro-preview sourceModelVersion Not reported thinkingEnabled Not reported ### Reported statistics Uncertainty Confidence interval: 28.06–29.88 contaminationWarning Not reported rank 3 reportedConfidenceHalfWidth 0.91 ### Source [Original source](https://labs.scale.com/leaderboard/vtb) Source scale Original model label gemini-3.1-pro-preview Observed 2026-10-21T16:35:36.362Z Evidence SHA-256 `70c39c91bb42d31fb0581b426c57134d54c74cc00ac658f5bd00694f52418c45` Source locator `/rows/2` |  |  |  |  |
| 3 | [Gemini 2.5 Pro Preview 06-05](https://benchize.com/models/gemini-2-5-pro-preview.md) | 11.75source score | CI 10.61–12.89 |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 11.75 \| source_score \| reported \| Higher is better \| ### Configuration Reasoning effort unknown Model configuration gemini-2.5-pro-preview-06-05 sourceModelVersion Not reported thinkingEnabled Not reported ### Reported statistics Uncertainty Confidence interval: 10.61–12.89 contaminationWarning Not reported rank 9 reportedConfidenceHalfWidth 1.14 ### Source [Original source](https://labs.scale.com/leaderboard/vtb) Source scale Original model label gemini-2.5-pro-preview-06-05 Observed 2026-10-21T16:35:36.362Z Evidence SHA-256 `70c39c91bb42d31fb0581b426c57134d54c74cc00ac658f5bd00694f52418c45` Source locator `/rows/8` |  |  |  |  |
| 4 | [Gemini 2.5 Flash](https://benchize.com/models/gemini-2-5-flash.md) | 4.69source score | CI 4.56–4.82 |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 4.69 \| source_score \| reported \| Higher is better \| ### Configuration Reasoning effort unknown Model configuration gemini-2.5-flash sourceModelVersion Not reported thinkingEnabled Not reported ### Reported statistics Uncertainty Confidence interval: 4.56–4.82 contaminationWarning Not reported rank 16 reportedConfidenceHalfWidth 0.13 ### Source [Original source](https://labs.scale.com/leaderboard/vtb) Source scale Original model label gemini-2.5-flash Observed 2026-10-21T16:35:36.362Z Evidence SHA-256 `70c39c91bb42d31fb0581b426c57134d54c74cc00ac658f5bd00694f52418c45` Source locator `/rows/15` |  |  |  |  |
| 5 | [Claude Sonnet 4](https://benchize.com/models/claude-sonnet-4.md) | 4.48source score | CI 4.07–4.89 |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 4.48 \| source_score \| reported \| Higher is better \| ### Configuration Reasoning effort unknown Model configuration claude-sonnet-4 sourceModelVersion Not reported thinkingEnabled Not reported ### Reported statistics Uncertainty Confidence interval: 4.07–4.89 contaminationWarning Not reported rank 17 reportedConfidenceHalfWidth 0.41 ### Source [Original source](https://labs.scale.com/leaderboard/vtb) Source scale Original model label claude-sonnet-4 Observed 2026-10-21T16:35:36.362Z Evidence SHA-256 `70c39c91bb42d31fb0581b426c57134d54c74cc00ac658f5bd00694f52418c45` Source locator `/rows/16` |  |  |  |  |

---
Cite as: VisualToolBench (VTB) — Benchize, https://benchize.com/benchmarks/scale%3Avtb (retrieved 2026-10-21)