Question 1

What is LTX-2?

Accepted Answer

LTX-2 is a family of video-generation model options. On this site, the hosted LTX-2.3 route supports text-to-video and image-to-video clips at 480p, 720p or 1080p for 5–20 seconds. LTX-2.5 is presented separately; check its provider documentation for release-specific features.

Question 2

Is LTX-2 free to use?

Accepted Answer

This site offers a free online entry with account credits. Availability, limits and pricing can change, so use the current pricing page as the source of truth.

Question 3

What is the difference between LTX-2.3, LTX-2.5 and 19B?

Accepted Answer

LTX-2.3 and LTX-2.5 are version labels for separate model routes. “19B” describes a model-size option used in some workflows; it is not a promise that every route exposes the same weights, license or hardware requirements.

Question 4

Who publishes LTX-2?

Accepted Answer

LTX model information should be verified against the official LTX organisation and the provider documentation linked in the guides on this site.

Question 5

What resolutions does the hosted LTX-2.3 route support?

Accepted Answer

The current hosted LTX-2.3 contract lists 480p, 720p and 1080p. Do not infer 2K or 4K support for this route from another release.

Question 6

How long can a hosted clip be?

Accepted Answer

The current contract accepts 5–20 second generation requests. Longer edits require multiple clips and post-production.

Question 7

Does the current hosted request include audio?

Accepted Answer

No native audio field is present in the current hosted text-to-video or image-to-video request. Add dialogue, ambience or music in a separate audio workflow unless a specific release documents otherwise.

Question 8

Which input modes are available?

Accepted Answer

Text-to-video and image-to-video are the two documented modes for the hosted route.

Question 9

What frame rate should I expect?

Accepted Answer

Frame-rate behavior depends on the selected provider workflow and output settings. The site does not promise a fixed 50 FPS figure.

Question 10

How do I start?

Accepted Answer

Open the free online page, sign in if required, choose a model and mode, then run a short prompt. The guides explain how to record the parameters for repeatable tests.

Question 11

How should I write a prompt?

Accepted Answer

Describe one shot: subject, action, camera, environment, lighting and mood. The prompts guide includes templates for text-to-video and image-to-video.

Question 12

What if a generation fails?

Accepted Answer

Retry with a shorter clip or lower resolution, then check the selected model and input. Credit handling follows the current account and pricing rules.

Question 13

Can I use outputs commercially?

Accepted Answer

Review the terms that apply to the model route and your account before commercial use. This site does not grant a blanket license for every LTX release.

Question 14

Can I add sound?

Accepted Answer

Yes, as a separate post-production step. The current hosted request should be treated as video-only unless the selected release explicitly documents audio output.

Question 15

What hardware is needed for local use?

Accepted Answer

Hardware requirements depend on the exact checkpoint and workflow. Follow the package documentation, check VRAM before selecting 720p or 1080p, and run a small 480p smoke test first.

Question 16

Does local setup work on Mac or AMD?

Accepted Answer

The hosted route is device-agnostic. Local acceleration depends on the package and backend; verify CUDA, ROCm or other support in the current provider documentation instead of assuming compatibility.

Question 17

Is ComfyUI required?

Accepted Answer

No. ComfyUI is an optional visual workflow tool. You can use another documented integration when the model package supports it.

Question 18

Can I fine-tune the model?

Accepted Answer

Only if the specific release and license document fine-tuning or adapters. This site makes no blanket LoRA or open-source promise.

Question 19

Where should I get model files?

Accepted Answer

Use the publisher or provider link listed by the workflow. Avoid unverified mirrors and record the exact file name and version.

Question 20

How many free credits are included?

Accepted Answer

The amount depends on the current account offer. Check the live pricing page rather than relying on a fixed number in an article.

Question 21

How are hosted generations priced?

Accepted Answer

Pricing is based on the current plan and generation settings. The pricing page lists the active credit schedule.

Question 22

Is an API available?

Accepted Answer

The site exposes documented API and studio routes for eligible accounts. Use the docs page for the current contract and authentication requirements.

Question 23

Where can I compare the model options?

Accepted Answer

Use the LTX-2.5 page and the comparison guide, then run the same prompt and duration before choosing a workflow.

LTX-2 AI FAQ

What is LTX-2?

Is LTX-2 free to use?

What is the difference between LTX-2.3, LTX-2.5 and 19B?

Who publishes LTX-2?

Start a documented test

Related resources