whatmodel.sh / video / HY-OmniWeaving
HY-OmniWeaving
↗ share
+ contribute
HY-OmniWeaving is an open-weights video model from Tencent, released March 2026. Its published size is 8.3B parameters. Licence: tencent-hy-community.
Catalogue identity notes
Tencent Hunyuan's OmniWeaving unified video generation/editing model on HunyuanVideo-1.5 (tencent/HY-OmniWeaving, pinned ddb4e6d; code released 3 Apr 2026, paper arXiv 2603.24458). Single public checkpoint. Rows: VideoArgus 'OmniWeaving' (8.3B), PhyGround 'OmniWeaving' (848x480, 81 frames, 16 fps, 50 steps), WROP 'HY-OmniWeaving' (open-weight, local). Licence: License.txt 'TENCENT HY COMMUNITY LICENSE AGREEMENT', 'Tencent HY-OmniWeaving Release Date: March 31, 2026'. Comparison-only: local route not verified.
Model information sources
Release date ↗ · Weights licence ↗ · Parameter count ↗
8.3B as printed in the VideoArgus Table 2 #P column (same backbone size as HunyuanVideo-1.5).
$ how it compares
The video models ranked either side of HY-OmniWeaving, out of 139 with a score, then more from Tencent.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 56 |  Kling 2.1 Master | 53±2 | api $0.28/sec | ||
| 57 |  Kandinsky 5.0 Pro | 52±2 | api $0.040/sec +local | ||
| 58 | HY-OmniWeaving | 52±4 | comparison only | ||
| 59 |  HunyuanVideo 1.5 | 52±2 | api $0.075/sec +local | ||
| 60 |  Cosmos3-Nano | 52±3 | local | ||
| more from Tencent | |||||
| 55 | HunyuanVideo 1.5 (I2V) | 53±2 | api $0.075/sec +local | ||
| 90 | HunyuanVideo-I2V | 45±4 | local | ||
| 105 | HunyuanVideo | 39±3 | api price unknown +local | ||
$ what it's good at
other tests4 not in a task · 93% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| WROP object-permanence exam (human Elo) | 1268.5 ✎ | ▄····██···▄█ beats 0/7 | 29% |
| VideoArgus-Bench T2V (rubric-grounded score) | 84.0% ✎ | █▄·▄█▄▄····▄ beats 4/8 | 25% |
| VideoArgus-Bench TI2V (rubric-grounded score) | 79.9% ✎ | █··▄▄▄▄····▄ beats 3/6 | 23% |
| PhyGround physics human evaluation (Overall, TI2V) | 62.0% ✎ | █▄▄··▄···▄·█ beats 5/7 | 17% |
$ where to run it
No verified API or download recorded yet · add one
Official weights and documentation: huggingface.co/tencent ↗
$ how the score was worked out · VideoArgus: Agentic Rubric-Gr… 48%, Training Object Permanence in… 29%, PhyGround: Benchmarking Physi… 17%
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 VideoArgus: Agentic Rubric-Grounded Unified Evaluation for Video Generation and Editing (arXiv:2608.05485v1) | 2 | 48% |
| Training Object Permanence in World Models (WROP exam; arXiv:2609.28654v1) | 1 | 29% |
| PhyGround: Benchmarking Physical Reasoning in Generative World Models (arXiv:2605.10806v1) | 1 | 17% |
| prior · pull toward the middle | 7% | |
| weighted average | = 52.4 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json