LTX 2 Fast
Neural Network
LTX 2 Fast is a Lightricks video model for fast generation: use it for drafts, previews, and short clips.
Specifications
- Input
- text, images
- Output
- video
- Video duration
- 6–20 s
- Resolution
- 1080p, 2k, 4k
- Audio generation
- Yes
Video generation
(duration 8 sec., 720p)
Video generation
(duration 8 sec., 1080p)
How it works LTX 2 Fast?
LTX 2 Fast is a fast version of LTX 2 by Lightricks, designed for scenarios where instant response is more important than long render times. The model is built for rapid brainstorming, real-time previews, and streaming video generation, including mobile use cases where you need clips immediately in multiple variations. In BotHub, it is accessible without a VPN or foreign cards; you can pay with local cards in rubles, and unused tokens do not expire. With over 250 neural networks in one window, you can create a draft here and send the final take to a heavy model without rewriting your integration—access is also available via a unified OpenAI-compatible API. It is useful for showing a client multiple scene variations in one evening, running a storyboard for an ad based on a text description, filling social media feeds with short videos, or estimating camera movement before expensive production. BotHub provides teams with contracts, invoices, EDI, and an admin panel with limits.Frequently asked questions about LTX 2 Fast
You can use generated results for commercial purposes. You own all rights to the content you create. The only restriction: make sure your prompt does not include copyrighted third-party material. You are responsible for respecting the rights to any input data.
ltx-2-fast is a generative video model by Lightricks. It creates short clips from text descriptions or animates uploaded images. Suitable for ad drafts, social media content, visualizing ideas, presentations, and quick scene drafts that can be refined later in editing.
The output is a video sequence, not text: the model renders a scene based on your request. Check the technical header and generation parameters for exact resolution and duration. The prompt context is 4095 tokens, with an output limit of 4096.
There is no specific reasoning mode noted in our data, and you shouldn't expect step-by-step text reasoning from a video model: its job is to turn descriptions into frames. Quality depends more on how detailed your description of the scene, lighting, and camera movement is.
There are no notes about function calling or structured JSON output in our data—the model's profile is purely visual. However, you can access it via the unified OpenAI-compatible API of BotHub, so integrating video generation into your service or pipeline is easy.
Images—yes: this is the image-to-video mode, where a static frame turns into a moving scene. It's convenient for animating product photos, illustrations, or concept art. There are no notes about accepting audio or video input in our data, so stick to text and images.
The BotHub interface, bot, and support are in Russian, payment is via ruble cards, and no VPN is needed. However, there is no information in our data on how the model reacts to Russian prompts, so describe the scene as detailed as possible and check the result with a short test.