✨ You asked, we delivered! To everyone running on an 8GB VRAM setup who thought they were locked out of top-tier video gen: this one is for you. 🤝
This advanced, two-stage workflow completely solves the budget VRAM crisis. It harnesses Minimax’s incredible motion adherence and base generation power as your foundation, then instantly hands off the baton to LTXV acting as a high-fidelity refiner.
The result? Crisp, high-resolution cinematic videos without triggering an Out-Of-Memory (OOM) error! 🤯
While this combo workflow has unmatched quality and fidelity, we've released a purely Minimax workflow. Check it out here. - https://civarchive.com/models/2941580/minimax-solo-int8-workflow-pure-8gb-power-low-storage-low-vram
V4 is in works. We're going to make this even better. READ THE CHANGELOG for list of Updates.
🛠️ Changelog & Bug Fixes
Known problem in V1.0: First few frames ignored by LTX Refiner , Artifacts produced in the Minimax stage cause Turbo LoRA's at the moment are not perfect.
Fixed in V2.0: Now your final output will have clarity and quality from the first frame itself. Also, LTX Refiner will fix the anatomy issues as well as any artifacts along with refining the video. Also, you can now use Interpolation along with RTX without much delay with Clean RAM VRAM nodes being added to the groups. A note is there in the workflow on how to connect RIFE Interpolation with RTX Upscaling.
Fixed in V3.0: Lower resolutions not refining well now give crystal clear output. Clarity has increased atleast 2-3x more. Added 2 loops for video extension. Unlimited loop video possibility. Switches given for streamlined rapid workflow as per requirements. Upscale Model stage added. T2V Mode is available now. Added another LoRA for LTX to refine videos. Get it from here - https://huggingface.co/CQdesign/LTX-2.5-CQ-Video-and-Image-Enhancer-LoRAs/blob/main/ltx2.5-CQ-enhancer-lora-for-videos-rank128.safetensors
🎨 The Feature Stack
8GB VRAM Friendly: Optimized explicitly for low-budget hardware using extreme INT4/INT8 quantization.
Two-Stage Refinement: Uses Minimax for raw motion framework, and LTXV to iron out details.
Smooth Frame Interpolation: 2x FPS motion smoothing baked right into the output line.
4K Upscaling Power: Blast your final render all the way up to sharp 4K resolution.
Backward Compatible: Don't have LTX 2.5? This pipeline is mostly backward-compatible with LTX 2.3 components!
🧠 The Shopping List: Required Models
To prevent errors, make sure you have these exact files placed in your ComfyUI directories:
⚡ Stage 1: The Minimax Engine
Checkpoint:
MinimaxH3INT8INT4_fl2vaINT8Pruned.safetensorsText Encoder:
Qwen3vl_32b_minimax_h3_int4_convrot.safetensorsVAEs:
minimax_h3_video_vae_fp16.safetensors&minimax_h3_audio_vae_fp32.safetensorsLoRA:
minimax_h3_turbo_v4_step600_ema_pruned_comfyui.safetensors
🌀 Stage 2: The LTXV Refiner & Upscaler (LTX 2.5 or 2.3)
VAEs:
LTX-2.5-audio-vae-bf16.safetensors<X-2.5-video-vae-conv-bf16.safetensorsTransformer:
LTX-2.5-22b-distilled-transformer-comfy-in8-convrot.safetensorsText Encoders:
Gemma4-12b-with-proj-ltx-2.5-comfyui-int8-convrot.safetensorsSpatial Upscaler:
LTX-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensorsLoRA:
LTX2.3_Crisp_Enhance.safetensors & ltx2.5-CQ-enhancer-lora-for-videos-rank128.safetensors
🛠️ Quick Start
Import the workflow file into your workspace.
Verify all paths point correctly to your quantized INT4/INT8 checkpoints.
Queue Prompt and enjoy elite video generation right on your local 8GB card!
💬 Join the Low VRAM Movement!
If this setup saves your GPU from bursting into flames, hit that Like button and leave a review!
👇 Low VRAM doesn't mean low quality! Drop your 4K interpolated video renders and test clips in the comments below! 👇
Description
FAQ
Comments (17)
Could you provide information about your gpu and production time? Thank you in advance for your efforts.
This is a turbo workflow with turbo LoRA. 4Step only. I use a Blackwell 5060 8GB.
For a 1024 x 768 243 frames 24fps 10 seconds with 2x Ultra Upscaling (No interpolation) - 330-350 seconds.
Please consider showing support by liking this workflow.
Wow. I couldn't have imagined this workflow working for me. Thankyou from bottom of my heart. ❤️
You're most welcome. Let me know if you've any other query.
Thanks. For all us poor stiffs with low ram gpus, :)
Haha. It's not our fault that they have amped the prices by 4-5x. Someone said, you need at least 12-16GB VRAM to run Minimax & my inner rebel said NO. If you can share your output here, it would be great. You're welcome.
Many thanks for your work. But...the upscaled video in the rapidflow workflow has only the half time of all others because the images for the video doesn`t come from the rife node and the fps are doubled. Is that a fault? I have amended it accordingly.
Hi. Much thanks for your comment, I went around & checked.
The math expression is inside RIFE group. If Interpolation is bypassed = math expression bypassed, it wont affect the fps. If you opted for 10 seconds, you ll get 10 seconds. Whether Interpolation is on or off. It might've happened if you pulled out the math expression from the group and kept the interpolation off. This is just to streamline the flow in case you want interpolation. Otherwise you can connect the fps in RTX straight from the FPS group.
Also I've added a note on how to connect RIFE with RTX. The issue is if one doesn't want to use RIFE, connecting the output from RIFE to RTX gives errors of input not found as RIFE is bypassed. So this is to get away from that. Now both can be used together without any delay in processing if you prefer to have that 2x FPS fidelity.
Also, I've updated both the workflows with Metadata. So you can download them again. I would suggest using Hyperstream. V3 is coming out soon.
So, please upload your outputs along with the CivitAI thumbnail here to get quick approval if you found this workflow to be working well for you.
@Mikel007 V3 is streamlined to avoid any of this. Let me know your input.
I don't know what type of black magic you did in this workflow, but it works!!!! Thankssss!!!!
So glad to know. And thanks for the buzz. V3 is coming soon & it'll have even more clarity and quality than V2. Share your outputs here please.
ran it default setting vram goes to 15.7gb, im on a 4080super 64gb ram, did i miss something ?
Have you turned off dynamic vram ? You ll need that. Though you have enough ram. Also run comfy with sage attention flag. If you are running into issues elsewhere too, I suggest using EZinstall(Easy Install) to help you with a clean install of Comfy. It rarely runs into a problem like the other self build. Let me know.
I'm a dumbass, love this workflow, keep up the good work my dude, but that firstframeselector stuff completely throws me for a loop, had to deactivate it
That's just to share your outputs here on Civit. Its meant to be uploaded with the video as the thumbnail so Civit is able to pull the metadata. Otherwise most posts don't pass their review unless they can check that it was indeed made from the workflow. It has no other business, so feel free to deactivate it if you want it that way. Thanks for the appreciation.