# OmniDocBench

> A document-understanding benchmark for parsing and reasoning over complex document layouts.

Canonical page: https://benchlm.ai/benchmarks/omnidocbench

- Category: [Multimodal & Grounded](/multimodal-grounded)
- Last updated: September 14, 2026

## About OmniDocBench

- Year: 2026
- Tasks: Complex document-understanding tasks
- Format: Document-understanding score
- Difficulty: Grounded document reasoning
- Paper: [Kimi K3: Open Frontier Intelligence](https://www.kimi.com/blog/kimi-k3)

The Kimi K3 launch table labels this row OmniDocBench without a version. BenchLM keeps it separate from the versioned OmniDocBench 1.5 key.

OmniDocBench is currently displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.

## Leaderboard (1 models)

| Rank | Model | Creator | Score |
|------|-------|---------|-------|
| 1 | [Kimi K3](/models/kimi-k3) | Moonshot AI | 91.1% |

## FAQ

### What does OmniDocBench measure?

A document-understanding benchmark for parsing and reasoning over complex document layouts.

### Which model scores highest on OmniDocBench?

Kimi K3 by Moonshot AI currently leads with a score of 91.1% on OmniDocBench.

### How many models are evaluated on OmniDocBench?

1 AI models have been evaluated on OmniDocBench on BenchLM.

### Does OmniDocBench affect BenchLM's overall score?

Not directly. OmniDocBench is still displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.