# AVA

AVA is an AI benchmark; this page records what it measures and traces its data, grader, and scores to primary sources.

Evaluates spatiotemporally localized recognition of 80 atomic actions in densely annotated movie clips.

## Details

- Address: Online, Global
- Website: [research.google.com/ava/](<https://research.google.com/ava/>)
- Tags: Science
- Category: Multimodal
- Status: Active
- Data licence: unknown
- Data access: open
- Grader: exact
- Organization: Google Research
- Released: 2017-05
- Paper: [arxiv.org/abs/1705.08421](<https://arxiv.org/abs/1705.08421>)
- Leaderboard: [research.google.com/ava/challenge.html](<https://research.google.com/ava/challenge.html>)
- Task count: 430
- Top-score provenance: unknown
- ABI family: No

## Questions

### Where is AVA?

AVA is at Online, Global.

### How do I contact AVA?

To contact AVA, visit https://research.google.com/ava/.

## Source

From [Benchmark primary sources](https://aibenchmarkindex.com/about), license CC0-1.0, checked 2026-10-15.

Last updated October 15, 2026.

## Links

- This benchmark on the [web page](<https://aibenchmarkindex.com/global/online/ava>)
- [Multimodal in Online](<https://aibenchmarkindex.com/global/online/category/multimodal.md>)
- [More benchmarks in Online](<https://aibenchmarkindex.com/global/online.md>)

> **Check before you compare.** Benchmark methods and leaderboards change. Follow the cited primary source before using a result in a decision.