whatmodel.sh / video / LTX-2.5
LTX-2.5
↗ share
+ contribute
LTX-2.5 is an open-weights video model from Lightricks, released August 2026. Its published size is 22B parameters. Licence: ltx-2.x-community.
$ how it compares
The video models ranked either side of LTX-2.5, out of 139 with a score, then more from Lightricks.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 51 |  Kling 2.1 Standard | 54±3 | api $0.050/sec | ||
| 52 |  Hailuo 02 | 54±2 | api $0.047/sec | ||
| 53 | LTX-2.5 | 53±6 | api $0.12/sec +local | ||
| 54 |  Vidu Q2 Pro | 53±3 | comparison only | ||
| 55 |  HunyuanVideo 1.5 (I2V) | 53±2 | api $0.075/sec +local | ||
| more from Lightricks | |||||
| 70 | LTX-2 | 49±2 | local | ||
| 83 | LTX-2 19B Distilled | 47±8 | api $0.020/sec +local | ||
| 112 | LTX-2.3 | 36±3 | api $0.040/sec +local | ||
$ what it's good at
text to video█····█·····█ #2 of 3 * · 49% of its score · 1 test
Trails Kandinsky 6.0 Video Lite here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| VABench T2AV micro Visual QA | 61.7% ✎ | ▄█··▄·▄···▄█ beats 0/7 | 49% |
other tests1 not in a task · 36% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Datapoint Video Bench (overall Elo) | 978.4 ✎ | ▃▃·▅·▃·▃··▃█ beats 2/9 | 36% |
$ where to run it
Through an API
api · Lightricks (LTX API) · $0.12/sec · docs.ltx.io ↗ · price sourceUSD per second of generated video, LTX API text-to-video, pro model at 1280x720. Product: LTX API v2/text-to-video. API model: ltx-2-5-pro. Settings: resolution=1280x720, tier=pro. Price basis: published cost per second. ltx-2-5-fast $0.09/s at 720p; pro 1080p $0.17/s, 1440p $0.25/s, 2160p $0.39/s. Currency: USD. LTX API pricing documentation, observed 2026-09-25.
Other ways to run it
version
- Run it as published · needs ~13.7 GB, NVIDIA only · from ltx.io
uv run python -m ltx_pipelines.distilled --transformer-path models/ltx-2.5/diffusion_models/ltx-2.5-22b-distilled-transformer-bf16.safetensors --text-encoder-path models/ltx-2.5/text_encoders/gemma4-12b-with-proj-ltx-2.5-bf16.safetensors --video-vae-path models/ltx-2.5/vae/ltx-2.5-video-vae-bf16.safetensors --audio-vae-path models/ltx-2.5/vae/ltx-2.5-audio-vae-bf16.safetensors --duration-head-path models/ltx-2.5/model_patches/ltx-2.5-duration-head-bf16.safetensors --spatial-upsampler-path models/ltx-2.5/latent_upscale_models/ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors --prompt "A golden retriever running through a sunny meadow, cinematic lighting" --seed 42 --output-path output_distilled.mp4 --quantization fp8-cast --offload cpu
Official weights and documentation: huggingface.co/Lightricks ↗
$ how the score was worked out · Kandinsky 6.0 Video: Foundati… 49%, Datapoint Video Bench leaderb… 36%
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 |
|---|---|---|
 Kandinsky 6.0 Video: Foundation Models for Synchronized Video and Audio Generation (arXiv:2610.05608v1) | 1 | 49% |
 Datapoint Video Bench leaderboard (datapointai/text-to-video-human-preferences-326k dataset card) | 1 | 36% |
| prior · pull toward the middle | 16% | |
| weighted average | = 53.1 |
$ 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