Blog
Latest posts from Upsampler.
LTX 2 Distilled: AI Video with Synchronized Audio
Generate AI video with synchronized audio using the LTX 2 family by Lightricks on Upsampler. One-time credits from 2 credits per second, no subscription.
Admin
@lucak5s
LTX 2 Distilled: AI Video with Synchronized Audio
Most AI video tools skip audio entirely - you get silent clips that need a separate soundtrack layer in post. LTX 2 Distilled, the open-source video model from Lightricks, generates synchronized audio and video in a single pass. On Upsampler it runs in the Video Generation tool from just 2 credits per second - one-time credits, no subscription.
Want to try AI video for free first? The Free AI Video Generator runs Wan 2.2 (silent video) with no signup, no watermark, and free daily GPU minutes.
What Is LTX 2 Distilled?
LTX is Lightricks' family of open-source video generation models. The Distilled variant is the efficient, low-step version - optimized to deliver the full LTX quality profile in fewer diffusion steps. Unlike most competing budget video models, LTX 2 natively generates audio alongside video, producing a finished clip in one go.
Highlights:
- Open-source video + audio generation
- Text-to-video, image-to-video, and audio generation in one model
- Versatile aspect ratios - 16:9, 9:16, 4:3, 3:4, 1:1, 21:9
- Synchronized audio output - ambient sound, effects, music
Why LTX 2 Distilled Matters
The Cheapest Path to Video with Audio
At 2 credits per second, LTX 2 Distilled is the most affordable audio-video model on Upsampler. If your use case depends on sound - atmospheric audio, sound effects, background music - this is the model to start with before stepping up to LTX 2.3 or premium models like Veo 3.1.
Text + Image + Audio in One Pass
Most workflows split video and audio generation into two tools: one generates silent video, another generates audio separately, and you composite them. LTX 2 Distilled does it all at once, which means:
- Audio matches the visuals - no sync issues
- No post-production compositing - the MP4 is ready to share
- Faster iteration - changing one prompt changes both the visual and the audio
Versatile Aspect Ratios
The model supports more aspect ratios than Wan 2.2 (which is mostly limited to 16:9 and 9:16), including:
- 16:9 - standard landscape
- 9:16 - portrait for TikTok, Reels, Shorts
- 4:3 - traditional TV / classic framing
- 1:1 - social media square
- 21:9 - ultra-widescreen cinematic
How to Use LTX 2 Distilled
- Open the Video Generation tool in the Upsampler dashboard.
- Select LTX 2 Distilled from the model picker.
- (Optional) Upload an image for image-to-video mode.
- Write a prompt describing the scene, motion, and the audio.
- Click generate.
- Download the MP4 (with audio).
A Cost-Effective Workflow
Because LTX generations cost credits per second, iterate on the visual composition for free first: the Free AI Video Generator runs Wan 2.2 5B with fast 4-step generation and no signup. When the composition works, re-run your best prompt through LTX 2 Distilled for the audiovisual final.
Prompting Tips for LTX 2 Distilled
Because the model generates audio as well as video, your prompts should describe the soundscape explicitly:
Describe the Visual
- Subject, environment, camera motion, lighting, style
Describe the Audio
- Ambient sound - "waves crashing", "city traffic", "forest birdsong", "rain on a metal roof"
- Sound effects - "footsteps on gravel", "door creaking open", "glass shattering"
- Music - "gentle piano melody", "pulsing electronic beat", "string quartet"
- Voice - "muffled conversation", "distant laughter"
Example Prompts
- "A coastal cliff at sunset, waves crashing below, seagulls circling, wide establishing shot, ambient ocean sounds and distant seagull cries"
- "Old library at night, candle flickers on a wooden desk, books stacked high, close-up shot, ambient room tone with quiet page-turning sounds"
- "Street busker playing acoustic guitar on a cobblestone plaza, people walking past, warm afternoon light, soft acoustic guitar melody, ambient crowd murmur"
- "Rain-slick Tokyo alley at night, neon signs, slow push-in, ambient rain sounds and distant traffic"
LTX 2 Distilled vs. the Free Video Models
| Model | Where | Mode | Audio | Best For |
|---|---|---|---|---|
| Wan 2.2 5B | Free tool | Text + Image | No | Quick iteration |
| Wan 2.2 14B | Free tool | Text + Image | No | High-quality animation |
| LTX 2 Distilled | Dashboard (2 cr/s) | Text + Image | Yes | Final audiovisual clips |
For the free video generator itself, head to the free AI video generator, no signup.
The Lightricks Video Model Family
The Upsampler dashboard includes the whole LTX family:
- LTX 2 Distilled (2 cr/s) - open-source audio-video, budget-friendly
- LTX 2.3 Fast (4 cr/s) - up to 20-second durations, the longest clips of any budget model
- LTX 2.3 Pro (6 cr/s) - highest quality LTX, near-premium output
No subscription required in the dashboard. One-time credits never expire - buy once and use whenever. See the budget video model comparison for deeper analysis.
When to Use LTX 2 Distilled
The entry-level LTX option is particularly well-suited for:
- Social media posts - an 8-second Reel or TikTok with matching audio
- Ambient background loops - website hero video with atmospheric sound
- Rapid audiovisual prototyping - test whether a concept works with both sight and sound before investing in premium tools
- Concept pitches - show clients what a scene will feel like, not just look like
For longer clips, higher resolutions, or professional-grade audio, step up to LTX 2.3 Fast or LTX 2.3 Pro in the dashboard.
Frequently Asked Questions
Is LTX 2 Distilled free on Upsampler?
LTX runs in the Video Generation tool at 2 credits per second - one-time credits, no subscription, credits never expire. To generate video for free, the Free AI Video Generator runs Wan 2.2 models (silent video) with free daily GPU minutes and no signup.
What is the cheapest way to make AI video with audio?
At 2 credits per second, LTX 2 Distilled is the cheapest audio-video model on Upsampler. A 5-second clip costs 10 credits with no monthly commitment. Iterate on the silent composition for free with Wan 2.2, then spend credits only on the final audiovisual version.
Does it really generate audio?
Yes. LTX 2 Distilled produces synchronized video + audio in a single pass. You can describe the soundscape in your prompt alongside the visual description.
What aspect ratios does it support?
LTX supports a wide range: 16:9, 9:16, 4:3, 3:4, 1:1, 21:9. Check the tool's dropdown for current options.
Can I use LTX 2 for image-to-video?
Yes. Upload an image and include motion + audio description in the prompt.
What if I need longer clips?
Free LTX 2 Distilled Turbo clips are relatively short. For up to 20 seconds per clip, use LTX 2.3 Fast in the Upsampler dashboard - no subscription, one-time credits never expire.
Can I use the videos commercially?
Yes. Check the underlying model's open-source license if you plan large-scale commercial distribution.
Make AI Video with Audio
LTX 2 Distilled delivers a complete audiovisual clip in a single generation at the lowest credit cost on the platform. For content that needs sound - social media, ambient loops, concept pieces - it is the right pick.
Open the Video Generation tool and select LTX 2 Distilled from the model picker - one-time credits, no subscription. Or start free with silent video on the Free AI Video Generator, no signup required.