Model card
Qwen2.5-VL 72B.
Alibabaopen-source72B paramsVision-Language ModelApache 2.01 current SOTA
§ 02 · Benchmarks
Every benchmark Qwen2.5-VL 72B has a recorded score for.
| # | Benchmark | Area · Task | Metric | Value | Rank | Date | Source |
|---|---|---|---|---|---|---|---|
| 01 | textvqa | — | accuracy | 85.5% | #1 | 2026-04-20 | source ↗ |
| 02 | ThaiOCRBench | Vision & Documents · Document OCR | ted-score | 0.7% | #5 | 2026-04-20 | |
| 03 | MME-VideoOCR | — | total-accuracy | 69.0% | #2 | 2026-04-20 | |
| 04 | OCRBench | — | score | 885.00 | #10 | 2025-02-19 | source ↗ |
| 05 | OmniDocBench | — | text-edit-distance | 0.2% | #2 | 2025-02-19 | source ↗ |
Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.
§ 03 · Strengths by area
Where Qwen2.5-VL 72B actually performs.
§ 06 · Sources & freshness
Where these numbers come from.
src
2
results
paperswithcode-public-api
2
results
codesota-api
1
result
3 of 5 rows marked verified. · first result 2025-02-19, latest 2026-04-20.