Benchmark Almanac

AGI-Eval Community Text to Video

doubao-seedance-2-0-260128 68.4 tops its leaderboard Wan 2.7 T2V 62.5 leads the models we track
Not in model cards

Models generate videos from text. The board reports objective, subjective and combined scores.

The 2026/11 leaderboard reports 30 scored model entries on its primary measure.

Released
February 2025
Kind
Video generation
Level
Everyday
Scored
Graded by people
Answers
Open-ended
Tasks
Kept private
Measures
Capability
In and out
Text → Video
Interaction
Single turn
Refresh
One fixed set
Run by
A host
Last model added
April 2026
Scores refreshed
November 21st, 2026
29
Entries on its own leaderboard, 8 models we track

doubao-seedance-2-0-260128 tops the leaderboard with 68.4 Wan 2.7 T2V leads with 62.5 among models released by April 2026

Best score per entry from the benchmark's own leaderboard, read on November 21st, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is Wan 2.7 T2V, at 62.5. This view shows only models we track, each at its best.

The newest model with a score here came out in April 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 doubao-seedance-2-0-260128 68.4
  2. 2 happyhorse-1.0-t2v 65.8
  3. 3 kling-v3 65.0
  4. 4 Wan 2.7 T2V 62.5
  5. 5 Veo-3.0-Generate-Preview 62.0
  6. 6 Vidu Q3 Pro 61.9
  7. 7 kling-v2-6 61.8
  8. 8 Wan 2.6 61.5
  9. 9 可灵2.5-Turbo 60.9
  10. 10 hailuo-2.3-t2v 60.4
  11. 11 Kling-2.0 60.2
  12. 12 Sora 2 60.0
  13. 13 Seedance-1.0-Pro 60.0
  14. 14 Hailuo-02 59.9
  15. 15 Seedance-1.5-Pro 59.8
  16. 16 Kling-2.1 59.5
  17. 17 Pixverse-V5 59.1
  18. 18 Wan2.2-T2V-A14B 59.0
  19. 19 API-SERVING 58.8
  20. 20 veo-3.1-generate-preview 58.5
  21. 21 Seedance-1.0 58.1
  22. 22 Sora 2 Pro 58.1
  23. 23 Pixverse-V4 57.8
  24. 24 Video-01 56.4
  25. 25 Vidu-Q1 56.2
  26. 26 Wan2.1-T2V-14B 55.5
  27. 27 Sora 54.0
  28. 28 Dreamina-S2.0-Pro 53.9
  29. 29 Pika-2.2 50.1

Source: Official AGI-Eval Community Text to Video leaderboard

#EntryScoreSettingReleasedRun onSource
1 doubao-seedance-2-0-260128 68.4 May 2026 Official leaderboard
2 happyhorse-1.0-t2v 65.8 May 2026 Official leaderboard
3 kling-v3 65.0 Mar 2026 Official leaderboard
4 Wan 2.7 T2V 62.5 Apr 2026 Apr 2026 Official leaderboard
5 Veo-3.0-Generate-Preview 62.0 Aug 2025 Official leaderboard
6 Vidu Q3 Pro 61.9 Feb 2026 Feb 2026 Official leaderboard
7 kling-v2-6 61.8 Dec 2025 Official leaderboard
8 Wan 2.6 61.5 Dec 2025 Dec 2025 Official leaderboard
9 可灵2.5-Turbo 60.9 Oct 2025 Official leaderboard
10 hailuo-2.3-t2v 60.4 Dec 2025 Official leaderboard
11 Kling-2.0 60.2 Jul 2025 Official leaderboard
12 Sora 2 60.0 Oct 2025 Oct 2025 Official leaderboard
13 Seedance-1.0-Pro 60.0 Jul 2025 Official leaderboard
14 Hailuo-02 59.9 Jul 2025 Official leaderboard
15 Seedance-1.5-Pro 59.8 Jan 2026 Official leaderboard
16 Kling-2.1 59.5 Jul 2025 Official leaderboard
17 Pixverse-V5 59.1 Sep 2025 Official leaderboard
18 Wan2.2-T2V-A14B 59.0 Jul 2025 Aug 2025 Official leaderboard
19 API-SERVING 58.8 Jan 2026 Official leaderboard
20 veo-3.1-generate-preview 58.5 Oct 2025 Official leaderboard
21 Seedance-1.0 58.1 Jul 2025 Official leaderboard
22 Sora 2 Pro 58.1 Nov 2025 Nov 2025 Official leaderboard
23 Pixverse-V4 57.8 Jul 2025 Official leaderboard
24 Video-01 56.4 Jul 2025 Official leaderboard
25 Vidu-Q1 56.2 Jul 2025 Official leaderboard
26 Wan2.1-T2V-14B 55.5 Feb 2025 Jul 2025 Official leaderboard
27 Sora 54.0 Aug 2025 Aug 2025 Official leaderboard
28 Dreamina-S2.0-Pro 53.9 Jul 2025 Official leaderboard
29 Pika-2.2 50.1 Jul 2025 Official leaderboard
#ModelScoreSettingReleasedRun onSource
1 Wan 2.7 T2V 62.5 Apr 2026 Apr 2026 Official leaderboard
2 Vidu Q3 Pro 61.9 Feb 2026 Feb 2026 Official leaderboard
3 Wan 2.6 61.5 Dec 2025 Dec 2025 Official leaderboard
4 Sora 2 60.0 Oct 2025 Oct 2025 Official leaderboard
5 Wan2.2-T2V-A14B 59.0 Jul 2025 Aug 2025 Official leaderboard
6 Sora 2 Pro 58.1 Nov 2025 Nov 2025 Official leaderboard
7 Wan2.1-T2V-14B 55.5 Feb 2025 Jul 2025 Official leaderboard
8 Sora 54.0 Aug 2025 Aug 2025 Official leaderboard

Rankings by part

We read 2 parts from the benchmark’s own leaderboard: Subjective accuracy and Objective accuracy. Each ranks models on one part of the test. Wan 2.7 T2V leads Subjective accuracy and Vidu Q3 Pro leads Objective accuracy. Arrows show how far a model moved from its overall rank among the models we track.

Subjective accuracy

8 models we track have a score here.

  1. 1 Wan 2.7 T2V 63.6
  2. 2 Vidu Q3 Pro 62.3
  3. 3 Wan 2.6 62.1
  4. 4 Sora 2 60.0
  5. 5 Wan2.2-T2V-A14B 58.6
  6. 6 Sora 2 Pro 57.0
  7. 7 Wan2.1-T2V-14B 53.5
  8. 8 Sora 51.3

6 models have held number one on the official leaderboard since its release in February 2025

Lab claims count from the day the card came out, independent runs from the model’s release.

Alibaba Group OpenAI Shengshu Technology
Own leaderboard Not tested since Apr 2026

Source: Official AGI-Eval Community Text to Video leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On the benchmark’s own leaderboard
52.5 55.0 57.5 60.0 62.5 2025 2026 52.5 55.0 57.5 60.0 62.5 2025 2026

Source: Official AGI-Eval Community Text to Video leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Audio and video

All 695 in Audio and video
  • Video-MMMU In use 300 expert-level lecture videos with questions testing perception, comprehension and adaptation of knowledge. Accuracy. Qwen 3.8 Max 88.7%
  • LVBench In use Hours-long video understanding with multiple-choice questions on reasoning, grounding and summarization. Gemini 4 Argon 91.7%
  • MMVU In use 3,000 expert-annotated questions on videos across 27 disciplines. Multiple-choice and open-ended accuracy. Qwen 3.8 Max 82.4%
  • MLVU In use Multi-task long video understanding (minutes to hours) with multiple-choice and generation tasks. Accuracy. Qwen 3.8 Max 90.8%
  • FLEURS In use Models transcribe read speech in 102 languages built from FLoRes-101 sentences. Model cards usually report word error rate, so lower scores are better. Phi-4-multimodal 4.00%
  • VoiceBench In use Spoken instructions with speaker, environment and content variation. Voice assistant responses scored by GPT judge or rules. Qwen3.5-Omni-Plus 93.1%

Data audit

8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 7 months ago, in April 2026, for Wan 2.7 T2V. Every benchmark

Catalog record 2 pass
  • Pass Task access The tasks are private.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
  • Pass Independent runs Scores from the official leaderboard (29 entries).
  • Pass Models scored 8 models we track, plus 21 other entries on its leaderboard.
  • Warning Newest score The newest came 7 months ago, in April 2026, for Wan 2.7 T2V.
  • Note Frontier models None of its models are on the capabilities index.
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on agi-eval.org: 8 rows for models we track and 21 other entries at the last read.
  • Pass Last read November 21st, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check