- 👁️ Pro-tier image-to-video model from Lightricks' LTX-2.5 generation
- 🧠 Lightricks describes a dual-stream diffusion transformer generating video and audio jointly
- 🆕 Successor to LTX-2.3; multi-shot scenes rendered in one pass
- 🔧 Per Lightricks, can edit existing footage and adapt to domains
- 💬 Outputs synchronized audio alongside the generated video
- ⚡ Pro targets final renders; the Fast variant targets quick iteration
- 📚 Open foundation model published by Lightricks, with a hosted API
- 🏢 Lightricks is the studio behind Facetune and LTX Studio
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 2.5 Pro is the quality-oriented image-to-video variant of Lightricks' LTX-2.5 generation, turning a still input frame into a clip with synchronized audio. Lightricks presents the LTX model as an open foundation system for cinematic audio–video creation, built on an asymmetric dual-stream diffusion transformer that handles both modalities in a single pass rather than bolting sound on afterwards.
Compared with its same-family predecessors — LTX Video 2.3 Full Quality and LTX Video 2.3 Fast — Lightricks positions 2.5 as the recommended starting point, citing multi-shot generation in one pass, the ability to edit existing footage, and domain adaptation as the main reasons to upgrade. The company also notes a migration caveat: custom LoRAs trained against the earlier latent space need retraining for 2.5.
Within the 2.5 release, Pro is the higher-fidelity path intended for final renders, while LTX Video 2.5 Fast covers previews and high-volume iteration across the same API. If your starting point is a written prompt rather than a reference image, LTX Video 2.5 Pro text-to-video handles that mode instead.
Because Lightricks publishes LTX weights openly alongside a hosted generation API, teams can either call the model as a service or run and fine-tune it on their own infrastructure.
This About section is AI-generated from public sources via VeniceStats + Venice inference, with no human editing. It may contain inaccuracies.
Daily usage · LTX Video 2.5 Pro · Organic sellers
| Seller | Health | Rep↓ | Sales | Video | Routing | $ / s | $ / video |
|---|---|---|---|---|---|---|---|
| Venice.ai Proxy 0x1f22…18c9 | failingupstream error · 3.7 h | 86.84 | $10.80 | 2.7 min | #1 | $0.0726 | — |
"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.