LTX Video 2.3 Fast
fal-ai/ltx-2.3/image-to-video/fastltx-2-v2-3-fast-image-to-video- - 🆕 Lightricks' open-source LTX-2.3 family release, dated March 2026.
- - 🎯 Image-to-video Fast variant optimized for iterative workflows.
- - 👁️ DiT-based foundation model with synchronized native audio.
- - 📏 Supports clips up to 20 seconds at resolutions reaching 4K.
- - 🔧 New VAE for sharper detail, plus LoRA fine-tuning.
- - 🔒 Released under Apache 2.0 with open weights on Hugging Face.
- - ⚡ Pairs with a Full Quality variant for final shots.
Lightricks is a creative technology company founded in January 2013, widely recognized for its popular selfie-editing app Facetune. Originally focused on mobile video and image editing tools, the company has expanded significantly into generative AI, developing its own video…
Explore 5 more models by Lightricks →LTX Video 2.3 Fast is the speed-optimized image-to-video endpoint within Lightricks' LTX-2.3 family, a diffusion-transformer (DiT) foundation model that generates synchronized video and audio from a single system. Per the catalog, the Fast tier is geared toward faster generation and iterative workflows. The family also ships a Full Quality tier: the LTX Video 2.3 Full Quality variant targets stable final shots, while a matching LTX Video 2.3 Fast text-to-video endpoint covers prompt-only generation.
Across the LTX-2.3 line, Lightricks' Hugging Face model card reports support for output up to 4K, clips as long as 20 seconds, native audio, a new VAE for sharper fine detail, and LoRA fine-tuning, all released with open weights under the Apache 2.0 license.
This release succeeds its same-family predecessor, LTX Video 2.0 Fast, the earlier speed-optimized image-to-video model in the LTX-2 line. The broader 19-billion-parameter LTX-2 checkpoints, including LTX Video 2.0 19B and its distilled counterpart, remain available for users who want the prior generation's weights. Training code, ComfyUI nodes, and reference workflows ship via the LTX-Video repository.
This About section is AI-generated from public sources via VeniceStats + Venice inference, with no human editing. It may contain inaccuracies.
| Seller | Rep↓ | Sales | Video | Routing | $ / s | $ / video |
|---|---|---|---|---|---|---|
| ▲ Apex Ant 0x73b4…e736 | 81.71 | — | — | #1 | $0.0296 | — |
| antseed-tidal-falcon-88cd 0x80d1…88cd | 12.78 | — | — | gated | $0.069 | — |
"Best price" and the seller table are live Antseed catalog data (advertised $/1M tokens — or $ per generated image for unit-billed image models — not settled amounts). Reputation = buyer trust score (0-100, the Antseed SDK's own formula). "Routing" = the SDK's default buyer routing (what the VPR desktop app ships with): a trust ≥ 60 gate on the effective reputation, then cheapest-first among routable sellers; live failover state (per-peer cooldowns) is buyer-side runtime and not included. Model knowledge (TLDR, provider, About) via the VeniceStats enrichment layer. Advertised catalog, not the model used in any specific purchase. "Usage on Antseed" counts only settlements whose buyers share the per-model split on-chain (metadata v2/v3, opt-in), so every usage figure is a lower bound.