JoyAI-Echo's performance on the LTX-2.5 dev transformer, with nothing baked on top. These are for people who want to apply their own distilled LoRA at their own strength - or run the model slow, as a dev model.Not few-step on their own. At 8 steps and cfg 1 these render soft and wrong. You either add a distilled LoRA (below) and run the normal distilled schedule, or run a dev recipe. If you just want to render, take the ready-made few-step files instead - Joy-LTX 2.5 on Civitai, or on Hugging Face <a target="_blank" rel="ugc" href="https://huggingface.co/joeygambino/joyai-echo-ltx25-echoVid-gguf">GGUF</a> / <a target="_blank" rel="ugc" href="https://huggingface.co/joeygambino/joyai-echo-ltx25-echoVid-comfy-native">comfy-native</a>. Those are exactly these dev merges with the official distilled LoRA baked in at 0.5.<h2 id="what-is-in-the-merge">What is in the merge</h2>JoyAI-Echo (a fine-tune of LTX-2.3) is the better actor - lip-sync, expression, a voice that stays put. LTX-2.5's transformer is shape-identical to 2.3's, so Echo's delta against 2.3-dev transplants cleanly onto 2.5-dev: 070T30 = 0.7 x the video attention / feed-forward delta, 0.3 x on the modulation tables (the default dose); 100T50 = 1.0 x / 0.5 x (livelier, hotter). Nothing retrained, no distillation applied - the file is dev + delta, full stop.<h2 id="how-to-use-them">How to use them</h2><ul><li>Your own distilled LoRA (the normal way). Put a Load LoRA (model only) node between the model loader and the sampler with the official <code>ltx-2.5-22b-distilled-lora-450-bf16.safetensors</code> (Lightricks/LTX-2.5 on Hugging Face, <code>loras/</code>), then run the distilled 8-step schedule at cfg 1. 0.4-0.6 is where the picture keeps its natural grade - measured on the same seed, 0.4 / 0.5 / 0.6 are indistinguishable by eye and all cleaner than 1.0; 1.0 is fully distilled and hotter. 0.5 is what the ready-made files use.</li><li>Any other distill / style LoRA the same way. That is the point of this page: the dial is yours.</li><li>As a dev model. 25-30 steps, the LTXV scheduler, cfg about 3. Slower and softer, and the Echo delta reads as extra guidance at cfg &gt; 1 - expect it hotter than stock dev; keep cfg modest.</li></ul>LoRA on the int8 file works in stock ComfyUI - it is how Lightricks ship LTX-2.5 dev + distilled LoRA. The <a target="_blank" rel="ugc" href="https://github.com/jlucasmcrell/ComfyUI-JoyLTX25">ComfyUI-JoyLTX25</a> canvases (Civitai: Joy-LTX 2.5 workflows) take these files too - drop the LoRA node after the loader.<h2 id="files-(one-version-per-dose)">Files (one version per dose)</h2><ul><li><code>LTX25dev-echoVid-&lt;dose&gt;-DiT-comfy-int8.safetensors</code> - 21.5 GB. Stock Load Diffusion Model; fastest on RTX 50, works on RTX 30/40 (ComfyUI 0.32+).</li><li><code>LTX25dev-echoVid-&lt;dose&gt;-DiT-comfy-fp8.safetensors</code> - 21.5 GB. comfy fp8_e4m3fn, scaled.</li><li><code>LTX25dev-echoVid-&lt;dose&gt;-DiT-bf16.safetensors</code> - 42.0 GB. The master, for merging or your own quantisation (if the upload here is refused for size it is on Hugging Face: <a target="_blank" rel="ugc" href="https://huggingface.co/joeygambino/joyai-echo-ltx25-echoVid-dev">joyai-echo-ltx25-echoVid-dev</a>).</li></ul>No GGUF on this page on purpose: a LoRA on a GGUF weight is dequantised per step and slow; quantise after you have baked your LoRA, or use the ready-made GGUF page.<h2 id="you-also-need-(lightricksltx-2.5-on-hugging-face)">You also need (Lightricks/LTX-2.5 on Hugging Face)</h2><ul><li><code>ltx-2.5-video-vae-bf16.safetensors</code> and <code>ltx-2.5-audio-vae-bf16.safetensors</code> → <code>models/vae/</code></li><li>text encoder <code>gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors</code> → <code>models/text_encoders/</code></li><li>the distilled LoRA above → <code>models/loras/</code>; the x2 latent upscaler if you run two passes.</li></ul><h2 id="why-this-exists">Why this exists</h2>The first Joy-LTX 2.5 build applied Echo's delta to the distilled 2.5 transformer and came out over-cooked - +16% saturation, hard contrast, blown highlights on any bright scene. Diagnosing it showed the delta behaves like extra guidance on a few-step model, and that a dev-base merge with the distill LoRA held back to 0.5 reads clean. That number suits my eye; these files let you pick yours.<h2 id="credits">Credits</h2>JoyAI-Echo by JD (jdopensource/JoyAI-Echo). LTX-2.5 and the distilled LoRA by Lightricks. Merge and files by me. Licensed under the LTX-2.x Community License (inherited from both parents). Questions: comment here or open a discussion on the Hugging Face page - I answer.

Joy-LTX 2.5 DEV merges bf16 / INT8 / fp8 - bring your own distilled LoRA

070T30 - the default dose - the LTX-2.5 dev transformer with JoyAI-Echo's video attention / feed-forward delta transplanted in, dosed at 0.7 on the video attention / feed-forward tensors and 0.3 on the modulation tables. Nothing is baked on top: these are not few-step files. You add the distillation, at the strength you want.Use it one of two ways<ul><li>Few-step (what most people want). Load <code>ltx-2.5-22b-distilled-lora-450-bf16.safetensors</code> (Lightricks/LTX-2.5 on Hugging Face, <code>loras/</code>) with a Load LoRA (model only) node between the model and the sampler, then run the normal distilled 8-step schedule at cfg 1. 0.4-0.6 keeps the picture's natural grade (0.5 is what the ready-made Joy-LTX 2.5 files use); 1.0 is fully distilled and hotter. Any other distill LoRA works the same way.</li><li>As a dev model. 25-30 steps, the LTXV scheduler, cfg about 3. Slower and softer - and the Echo delta reads as extra guidance above cfg 1, so keep cfg modest or it warms up fast.</li></ul>The three files<ul><li><code>LTX25dev-echoVid-070T30-DiT-comfy-int8.safetensors</code> - 21.5 GB. Stock Load Diffusion Model; fastest on RTX 50, fine on RTX 30/40 (ComfyUI 0.32+). Start here.</li><li><code>LTX25dev-echoVid-070T30-DiT-comfy-fp8.safetensors</code> - 21.5 GB. comfy fp8_e4m3fn, scaled - the alternative if int8 misbehaves on your card.</li><li><code>LTX25dev-echoVid-070T30-DiT-bf16.safetensors</code> - 42.0 GB. The master: for merging, for your own quantisation, or a 48 GB+ card. LoRA on bf16 is the cleanest of the three.</li></ul>This is the dose the released few-step files are built from and the one to start with: cleaner skin on quiet scenes, close-ups and interviews, and the most natural grade of the two.Also needed from <a target="_blank" rel="ugc" href="https://huggingface.co/Lightricks/LTX-2.5">Lightricks/LTX-2.5</a>: both VAEs, the Gemma-4 text encoder, and the distilled LoRA above. Workflows: <a target="_blank" rel="ugc" href="https://github.com/jlucasmcrell/ComfyUI-JoyLTX25">ComfyUI-JoyLTX25</a> (drop a LoRA node after the loader - the PLUS canvases already have a four-slot LoRA stack there). Ready-made few-step files: Joy-LTX 2.5 on Civitai, or <a target="_blank" rel="ugc" href="https://huggingface.co/joeygambino/joyai-echo-ltx25-echoVid-gguf">GGUF</a> / <a target="_blank" rel="ugc" href="https://huggingface.co/joeygambino/joyai-echo-ltx25-echoVid-comfy-native">comfy-native</a> on Hugging

Files with SHA256: 7a1c31bf99156b0d3c5de527eee17f49d346085980c23f6e5c4c38e965307f17