# PerceptionBench (tools) leaderboard

Multimodal & Grounded benchmark

> Every figure below is reproduced as its upstream source published it: nothing is modelled, estimated, interpolated or converted. `Unavailable` means no source published the value — it is never a zero. Each value carries its evidence state and the date it was observed.

Page: https://modelchoice.dev/leaderboards/benchmarks/multimodal-grounded-perception-bench-tools

PerceptionBench with tools. Every model the catalog carries a published PerceptionBench (tools) value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Multimodal & Grounded |
| Measure | Provider-reported score |
| Tasks | Fine-grained visual-perception tasks with tools |
| Difficulty | Fine-grained visual perception |
| Published by | [Introducing Muse Ember](https://research.meta.ai/blog/muse-ember) |

PerceptionBench fine-grained visual perception tasks run with tool use, as reported by Meta.

## PerceptionBench (tools) ranking

3 models with a published PerceptionBench (tools) value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.

| Rank | Model | Provider | Provider-reported score | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [Muse Ember](https://modelchoice.dev/models/muse-ember) | Meta | 80.6 | Observed 2026-11-21 · source benchlm:models |
| 2 | [Claude Haiku 5.5](https://modelchoice.dev/models/claude-haiku-5-5) | Anthropic | 78.6 | Observed 2026-11-21 · source benchlm:models |
| 3 | [GPT-6 Luna](https://modelchoice.dev/models/gpt-6-luna) | OpenAI | 62.8 | Observed 2026-11-21 · source benchlm:models |

Rows are ordered by the value the source published, highest first. The source does not state whether a higher value is the better result, so this document does not either: for a benchmark that measures a rate of failure, read the table from the bottom.
