TencentVideo model
HunyuanVideo 1.5
In image to video, HunyuanVideo 1.5 is in place 18 of 21, with text only in place 20.
1,198Arena ratinghigher is better
1,169Arena ratinghigher is better
n/alongest single cliphigher is better
30out of 100same as last month
Summary
New in HunyuanVideo 1.5 is Tencent’s compact 8.3-billion-parameter open-weight approach to text-to-video and image-to-video generation. Base outputs are 480p or 720p; a separate video super-resolution module scales them to 1080p. Native audio is absent.
The architecture combines a DiT with Selective and Sliding Tile Attention to reduce the cost of longer video sequences. Step-distilled checkpoints accelerate image-to-video generation. The 1080p result is therefore a super-resolution output, not the model’s native resolution.
HunyuanVideo 1.5 benchmarks
Image to video arena
How people rate generated videos in blind comparisons, Elo points · higher is better
20 of 21 models
as of September 22, 2026 · Source: LMArena
Text to video arena
How people rate generated videos in blind comparisons, Elo points · higher is better
20 of 23 models
as of September 22, 2026 · Source: LMArena
Show all 154 individual results from 22 sources
Page 1 of 13
Published benchmarks
LMArena Text to Video
Elo ratings summarize pairwise comparisons within this leaderboard. They are not percentages. The scale does not express absolute quality or linear gaps between models.
higher is betterAs of November 5, 2026Scale: 0 Elo - 1,516 Elo
Methodology and context
- Metric
- elo-rating
- Scale
- No upper bound is documented. The chart spans the recorded values and zero.
- Participants
- 23
- Sample
- 4,292
LMArena Image to Video
Elo ratings summarize pairwise comparisons within this leaderboard. They are not percentages. The scale does not express absolute quality or linear gaps between models.
higher is betterAs of November 5, 2026Scale: 0 Elo - 1,495 Elo
Methodology and context
- Metric
- elo-rating
- Scale
- No upper bound is documented. The chart spans the recorded values and zero.
- Participants
- 21
- Sample
- 5,462
Measurements without matching peer values
These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.
WorldReasonBench (Show measurement, test conditions, and source)
- Source value
- 20.2
- Score
- 20.2
- Metric
- AccQA (question-answer accuracy)
- Unit
- %
- Benchmark version
- arXiv:2605.10434v1, Appendix Table 37 (overall)
- Category
- video-world-reasoning
- Direction
- Higher is better
- Sample
- 80
- Lower confidence bound
- 16
- Upper confidence bound
- 24.6
- Source type
- independent-evaluator
- Evaluator
- WorldReasonBench
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. WorldReasonBench shared 80-item evaluation cohort; exact HunyuanVideo 1.5 row. Scores are reported on a 0-100 percentage scale with 95% bootstrap confidence intervals.
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
WorldReasonBench (Show measurement, test conditions, and source)
- Source value
- 17.9
- Score
- 17.9
- Metric
- Score_PR (process-reasoning score)
- Unit
- %
- Benchmark version
- arXiv:2605.10434v1, Appendix Table 37 (overall)
- Category
- video-world-reasoning
- Direction
- Higher is better
- Sample
- 80
- Lower confidence bound
- 13.8
- Upper confidence bound
- 22.4
- Source type
- independent-evaluator
- Evaluator
- WorldReasonBench
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. WorldReasonBench shared 80-item evaluation cohort; exact HunyuanVideo 1.5 row. Scores are reported on a 0-100 percentage scale with 95% bootstrap confidence intervals.
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
WorldReasonBench (Show measurement, test conditions, and source)
- Source value
- 27
- Score
- 27
- Metric
- S(v) (weighted human rating)
- Unit
- %
- Benchmark version
- arXiv:2605.10434v1, Appendix Table 37 (overall)
- Category
- video-world-reasoning
- Direction
- Higher is better
- Sample
- 80
- Lower confidence bound
- 21.9
- Upper confidence bound
- 33.3
- Source type
- independent-evaluator
- Evaluator
- WorldReasonBench
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. WorldReasonBench shared 80-item evaluation cohort; exact HunyuanVideo 1.5 row. Scores are reported on a 0-100 percentage scale with 95% bootstrap confidence intervals.
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
RISE-Video (Show measurement, test conditions, and source)
- Source value
- 38.1
- Score
- 38.1
- Metric
- Reasoning Alignment (RA)
- Unit
- %
- Benchmark version
- arXiv:2602.05986v1, Table 1
- Category
- text-image-to-video
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- RISE-Video
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. RISE-Video Table 1 row HunyuanVideo-1.5-720P-I2V; preserve this exact 720p image-to-video variant. The paper does not provide a per-model denominator for these metrics.
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
RISE-Video (Show measurement, test conditions, and source)
- Source value
- 75
- Score
- 75
- Metric
- Temporal Consistency (TC)
- Unit
- %
- Benchmark version
- arXiv:2602.05986v1, Table 1
- Category
- text-image-to-video
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- RISE-Video
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. RISE-Video Table 1 row HunyuanVideo-1.5-720P-I2V; preserve this exact 720p image-to-video variant. The paper does not provide a per-model denominator for these metrics.
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
RISE-Video (Show measurement, test conditions, and source)
- Source value
- 68.4
- Score
- 68.4
- Metric
- Physical Rationality (PR)
- Unit
- %
- Benchmark version
- arXiv:2602.05986v1, Table 1
- Category
- text-image-to-video
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- RISE-Video
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. RISE-Video Table 1 row HunyuanVideo-1.5-720P-I2V; preserve this exact 720p image-to-video variant. The paper does not provide a per-model denominator for these metrics.
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
RISE-Video (Show measurement, test conditions, and source)
- Source value
- 92.6
- Score
- 92.6
- Metric
- Visual Quality (VQ)
- Unit
- %
- Benchmark version
- arXiv:2602.05986v1, Table 1
- Category
- text-image-to-video
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- RISE-Video
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. RISE-Video Table 1 row HunyuanVideo-1.5-720P-I2V; preserve this exact 720p image-to-video variant. The paper does not provide a per-model denominator for these metrics.
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
RISE-Video (Show measurement, test conditions, and source)
- Source value
- 60.4
- Score
- 60.4
- Metric
- Weighted overall score (W.Score)
- Unit
- %
- Benchmark version
- arXiv:2602.05986v1, Table 1
- Category
- text-image-to-video
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- RISE-Video
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. RISE-Video Table 1 row HunyuanVideo-1.5-720P-I2V; preserve this exact 720p image-to-video variant. The paper does not provide a per-model denominator for these metrics.
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
RISE-Video (Show measurement, test conditions, and source)
- Source value
- 8.6
- Score
- 8.6
- Metric
- Accuracy
- Unit
- %
- Benchmark version
- arXiv:2602.05986v1, Table 1
- Category
- text-image-to-video
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- RISE-Video
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. RISE-Video Table 1 row HunyuanVideo-1.5-720P-I2V; preserve this exact 720p image-to-video variant. The paper does not provide a per-model denominator for these metrics.
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
VIPER (Show measurement, test conditions, and source)
- Source value
- 8.4
- Score
- 8.4
- Metric
- POC@1.0, temporal
- Unit
- %
- Benchmark version
- ACL 2026 long paper, Table 2
- Category
- temporal
- Direction
- Higher is better
- Sample
- 43
- Source type
- independent-evaluator
- Evaluator
- VIPER (ACL 2026)
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. VIPER Table 2 image-conditioned video reasoning, one generated video per prompt, 309 items overall; domain rows retain their published domain counts. Main metric is POC@1.0 with GPT-5 judging; OC and outcome-hacking rate are reported separately.
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
Recorded measurement details
LMArena Text to Video: LMArena Text to Video (Show measurement, test conditions, and source)
- Source value
- 1169
- Score
- 1169
- Metric
- elo-rating
- Unit
- Elo
- Task
- lmarena-text-to-video
- Task label
- LMArena Text to Video
- Direction
- Higher is better
- Source rank
- 38
- Sample
- 4292
- Lower confidence bound
- 1153
- Upper confidence bound
- 1185
- Retrieved at
- 2026-11-05
LMArena Image to Video: LMArena Image to Video (Show measurement, test conditions, and source)
- Source value
- 1198
- Score
- 1198
- Metric
- elo-rating
- Unit
- Elo
- Task
- lmarena-image-to-video
- Task label
- LMArena Image to Video
- Direction
- Higher is better
- Source rank
- 40
- Sample
- 5462
- Lower confidence bound
- 1182
- Upper confidence bound
- 1214
- Retrieved at
- 2026-11-05
Page 1 of 13
Video sample from our test
This video exists as a documented run for the exact model version. Playback is muted.
Motion and physical consistency
This sample tests Anatomy, object consistency, camera tracking, rain, and reflections.
Prompt and settings
Generate exactly one continuous five-second 16:9 cinematic video. A fictional bicycle courier in a mustard-yellow raincoat rides a red city bicycle from left to right across a rain-wet Hamburg street at blue hour for the entire clip. The camera tracks her smoothly at waist height. Both hands remain visible on the handlebars, the bicycle geometry stays physically correct, and a transparent umbrella fixed to the bicycle moves naturally in the wind. Wet-street reflections and raindrops follow realistic physics. No cuts, no slow motion, no text, no logos, no subtitles, and no audio.
Generated through fal.ai with model fal-ai/hunyuan-video-v1.5/text-to-video. Archived settings were 16:9, one request, first output, requested 5 s, seed 1928510983, request {"prompt":"Generate exactly one continuous five-second 16:9 cinematic video. A fictional bicycle courier in a mustard-yellow raincoat rides a red city bicycle from left to right across a rain-wet Hamburg street at blue hour for the entire clip. The camera tracks her smoothly at waist height. Both hands remain visible on the handlebars, the bicycle geometry stays physically correct, and a transparent umbrella fixed to the bicycle moves naturally in the wind. Wet-street reflections and raindrops follow realistic physics. No cuts, no slow motion, no text, no logos, no subtitles, and no audio.","aspect_ratio":"16:9","resolution":"480p","num_frames":121,"num_inference_steps":28,"enable_prompt_expansion":false}, configured price estimate $0.38.
Popularity compared
How often the model is searched for, used and mentioned, compared with every model of its category.
Popularity index
0 to 100, higher is more popular
20 of 98 models
As of November 2026 · Source: gradually.ai popularity index
Over time
Weekly values from October 18, 2026 to November 5, 2026
HunyuanVideo 1.5Dreamina Seedance 2.0 720pWan 2.7Kling 2.6 ProVeo 3.1 Fast
HunyuanVideo 1.5 compared
Compare with another model
The most popular comparisons
- 1Gemini Omni Flash vs. HunyuanVideo 1.5
- 2FLUX.3 Video vs. HunyuanVideo 1.5
- 3Dreamina Seedance 2.5 720p vs. HunyuanVideo 1.5
- 4Dreamina Seedance 2.0 720p vs. HunyuanVideo 1.5
- 5MiniMax H3 vs. HunyuanVideo 1.5
- 6Sora 2 Pro vs. HunyuanVideo 1.5
- 7Sora 2 vs. HunyuanVideo 1.5
- 8Veo 3.1 vs. HunyuanVideo 1.5
- 9Veo 3.1 Fast vs. HunyuanVideo 1.5
Show all 48 comparisons
- 10Grok Imagine Video 1.5 vs. HunyuanVideo 1.5
- 11Grok Imagine Video vs. HunyuanVideo 1.5
- 12Wan 2.7 vs. HunyuanVideo 1.5
- 13Wan 2.6 vs. HunyuanVideo 1.5
- 14Runway Gen-4.5 vs. HunyuanVideo 1.5
- 15Kling 3.0 Pro vs. HunyuanVideo 1.5
- 16Kling 2.6 Pro vs. HunyuanVideo 1.5
- 17Vidu Q3 Pro vs. HunyuanVideo 1.5
- 18PixVerse V5.6 vs. HunyuanVideo 1.5
- 19Hailuo 2.3 vs. HunyuanVideo 1.5
- 20LTX-2 19B vs. HunyuanVideo 1.5
- 21Wan 2.2 A14B vs. HunyuanVideo 1.5
- 22HunyuanVideo 1.5 vs. Pika 2.2
- 23HunyuanVideo 1.5 vs. Gemini Omni 1.1 Flash
- 24HunyuanVideo 1.5 vs. Muse Video
- 25HunyuanVideo 1.5 vs. Wan 3.0
- 26HunyuanVideo 1.5 vs. Wan 3.0 Prime
- 27HunyuanVideo 1.5 vs. HappyHorse 1.0
- 28HunyuanVideo 1.5 vs. HappyHorse 1.1
- 29HunyuanVideo 1.5 vs. PixVerse V6
- 30HunyuanVideo 1.5 vs. PixVerse C1
- 31HunyuanVideo 1.5 vs. Seedance 2.0 Fast
- 32HunyuanVideo 1.5 vs. Seedance 2.0 Mini
- 33HunyuanVideo 1.5 vs. Vidu Q3 Turbo
- 34HunyuanVideo 1.5 vs. Vidu Q2
- 35HunyuanVideo 1.5 vs. Vidu Q1
- 36HunyuanVideo 1.5 vs. Kling 4.0
- 37HunyuanVideo 1.5 vs. Kling 3.0 Standard API
- 38HunyuanVideo 1.5 vs. Kling Video O1
- 39HunyuanVideo 1.5 vs. Hailuo 2.3 Fast
- 40HunyuanVideo 1.5 vs. Wan 2.5 Preview
- 41HunyuanVideo 1.5 vs. Wan 2.6 I2V Flash
- 42HunyuanVideo 1.5 vs. Wan 2.2 TI2V 5B
- 43HunyuanVideo 1.5 vs. Wan 2.2 I2V A14B
- 44HunyuanVideo 1.5 vs. HunyuanVideo 1.0 T2V
- 45HunyuanVideo 1.5 vs. HunyuanVideo I2V 1.0
- 46HunyuanVideo 1.5 vs. Ray 3.2
- 47HunyuanVideo 1.5 vs. Grok Imagine Video 1.5 Lite
- 48HunyuanVideo 1.5 vs. LTX-2.5
All specifications
Architecture
Input and output
Watermarking and provenance
AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.
- AI watermarking
- Unknown
- no source
- Watermarking technique
- Not known
- no source
Licence and access
Prices
- fal.ai (fal-ai/hunyuan-video-v1.5/text-to-video)
- $0.38 per request (estimated)Source
Frequently asked questions about HunyuanVideo 1.5
How good is HunyuanVideo 1.5?
Rank 20 of 23 in the text-to-video arena rating. With a start image: rank 18 of 21.
LMArena, as of Sep 22, 2026
Sources and data date
Show all 35 sources
Benchmark sources at a glance
Which independent evaluators tested this model. This is coverage, not a quality ranking.
| Evaluator | Test series | Composite indices | Recorded active measurements |
|---|---|---|---|
| LMArena | 0 | 0 | 2 |
Without vendor claims and historical or disputed results; those stay labelled in the single values. 2 measurements are not yet assigned to a test series.
| Type | Evidence and data date |
|---|---|
| Research status | Research date October 16, 2026. 31 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Not found in the inspected sources | arena-video-edit-exact-row Note The inspected Video Edit table has no exact model-version row. T2V and I2V values were not reused. |
| Not found in the inspected sources | vabench-exact-model-row Note VABench has no exact model row. Sora 2 was not transferred to Sora 2 Pro or other models. |
| Not found in the inspected sources | artificial-analysis-t2v-exact-row Note No exact model version or resolution was found in the inspected Artificial Analysis table. |
| Not found in the inspected sources | artificial-analysis-i2v-exact-row Note No exact model version or resolution was found in the inspected Artificial Analysis table. |
| Unresolved | vbench-current-public-leaderboard Note The current VBench leaderboard is embedded in a dynamic Hugging Face iframe and was not retrievable; current scores remain unresolved. |
| Not found in the inspected sources | vidbenchmark-exact-row Note No exact model id is present in the 2026-08-04 dataset; similarly named variants were not transferred. |
| Unresolved | hash-pinned-checkpoint-and-generation-seed Original wording OSCBench names the evaluated system HunyuanVideo-1.5 (Team, 2025) and gives official/default implementation settings, but no weights revision hash or generation seed; none is inferred. |
| Not published | confidence-intervals-and-human-rater-dispersion Original wording The cited result tables give point scores without per-cell confidence intervals or per-rater dispersion; none is inferred. |
| Additional source | Tencent · Tencent HunyuanVideo 1.5 model card · Editorial description reviewed October 2, 2026 |
| Additional source | fal.ai |
| Additional source | LMArena (retrieved November 5, 2026)12 |
| Additional source | VGI-Bench, arXiv:2608.19583v3 (retrieved October 16, 2026) |
| Additional source | Physics-IQ Verified GitHub leaderboard (retrieved October 16, 2026) |
| Additional source | H2R-Bench, arXiv:2608.13049 (retrieved October 16, 2026) |
| Additional source | WorldReasonBench (retrieved October 16, 2026) |
| Additional source | RISE-Video (retrieved October 16, 2026) |
| Additional source | VIPER (ACL 2026) (retrieved October 16, 2026) |
| Additional source | Video Models by Wiro AI (retrieved October 16, 2026; October 28, 2026) |
| Additional source | Han et al., OSCBench: Benchmarking Object State Change in Text-to-Video Generation, ACL 2026 (retrieved October 17, 2026) |
| Additional source | TextVBench: Evaluating Text Rendering and Preservation in Video Generation (retrieved October 17, 2026) |
| Additional source | WorldGuide: Goal-Directed Video World Model for Procedural Task Execution (retrieved October 17, 2026) |
| Additional source | VTR-Bench, Table 1 (retrieved October 17, 2026; October 19, 2026) |
| Additional source | VGI-Bench paper (retrieved October 17, 2026) |
| Additional source | Sci-VBench paper, arXiv:2608.09873v1 (retrieved October 17, 2026; October 23, 2026) |
| Additional source | Physics-IQ Verified paper, arXiv:2606.18943 (retrieved October 17, 2026) |
| Additional source | Dynamic Video Generation: Shaping Video Generation Across Time and Space (arXiv:2605.21042v1) (retrieved October 20, 2026) |
| Additional source | SpecLoR: Spectral Lookahead Rectification for Motion-Coherent Text-to-Video Generation (arXiv:2606.11969) (retrieved October 21, 2026) |
| Additional source | RoboTrustBench (retrieved October 23, 2026) |
| Additional source | WFM-Eval (retrieved October 23, 2026) |
| Additional source | Arena Video Image-to-Video (retrieved September 21, 2026; October 16, 2026; October 28, 2026) |
| Additional source | Arena Video Text-to-Video (retrieved September 21, 2026; October 16, 2026; October 28, 2026) |
| Additional source | Model metadata (retrieved October 16, 2026)1234 |
| Additional source | Model metadata (retrieved October 16, 2026; October 28, 2026) |
| Additional source | OpenRouter: Popularity index (derived), CC BY 4.0 |
| Additional source | Vercel AI Gateway: Popularity index (derived), CC BY 4.0 |
| Additional source | Ollama: cumulative downloads for the popularity index |
| Additional source | Replicate: cumulative runs for the popularity index |
| Additional source | Google Trending Now: limited feed selection for Germany and the US |
| Additional source | Creative Commons Attribution 4.0 |