Home › AI Workflows › 60-second AI short film, narrated 60-second AI short film, narrated The whole thing: script, keyframes, clips and a voiceover. Every stage a measured first-party number, nothing assumed.
The pipeline Write the script and shot list , qwen3-32bGenerate 15 keyframes , stable-diffusion-xlRender 15 clips (~4s each) , ltx-videoNarrate 60 seconds of voiceover , kokoro-ttsFastest card where every stage is a real measurement: NVIDIA H100 80GB HBM3 , 2.6 min for the whole job.
Every GPU, slowest job to fastest Cards that can't run this pipeline Every one of these fails on the same kind of wall, a stage that will not fit in VRAM.
AMD Radeon RX 6700 , blocked at "Write the script and shot list" (Qwen3 32B)AMD Radeon RX 6800 , blocked at "Write the script and shot list" (Qwen3 32B)AMD Radeon RX 6800 XT , blocked at "Write the script and shot list" (Qwen3 32B)AMD Radeon RX 6900 XT , blocked at "Write the script and shot list" (Qwen3 32B)AMD Radeon RX 6950 XT , blocked at "Write the script and shot list" (Qwen3 32B)AMD Radeon RX 7600 , blocked at "Write the script and shot list" (Qwen3 32B)AMD Radeon RX 7700 XT , blocked at "Write the script and shot list" (Qwen3 32B)AMD Radeon RX 7800 XT , blocked at "Write the script and shot list" (Qwen3 32B)AMD Radeon RX 9070 , blocked at "Write the script and shot list" (Qwen3 32B)AMD Radeon RX 9070 XT , blocked at "Write the script and shot list" (Qwen3 32B)GeForce GTX 1080 Ti , blocked at "Write the script and shot list" (Qwen3 32B)GeForce RTX 4060 , blocked at "Write the script and shot list" (Qwen3 32B)GeForce RTX 4080 Super , blocked at "Write the script and shot list" (Qwen3 32B)GeForce RTX 5060 Ti , blocked at "Write the script and shot list" (Qwen3 32B)GeForce RTX 5070 , blocked at "Write the script and shot list" (Qwen3 32B)GeForce RTX 5070 Ti , blocked at "Write the script and shot list" (Qwen3 32B)GeForce RTX 5080 , blocked at "Write the script and shot list" (Qwen3 32B)Intel Arc A750 , blocked at "Write the script and shot list" (Qwen3 32B)Intel Arc A770 Limited Edition , blocked at "Write the script and shot list" (Qwen3 32B)Intel Arc B580 , blocked at "Write the script and shot list" (Qwen3 32B)Intel Arc Pro A60 , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce GTX 1070 Ti , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce GTX 1080 , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce GTX 1660 , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce GTX 1660 Super , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce GTX 1660 Ti , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce RTX 2060 , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce RTX 2060 Super , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce RTX 2070 , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce RTX 2070 SUPER , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce RTX 2080 Founders Edition , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce RTX 2080 Super , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce RTX 2080 Ti Founders Edition , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce RTX 3050 , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce RTX 3060 , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce RTX 3060 Ti , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce RTX 3070 Founders Edition , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce RTX 3070 Ti , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce RTX 3080 , blocked at "Write the script and shot list" (Qwen3 32B)NVIDIA GeForce RTX 3080 Ti , blocked at "Write the script and shot list" (Qwen3 32B)…and 18 more.
How these numbers are built Each stage time is the quantity of work divided by that card's measured throughput for that model, from our own bench. The pipeline is assumed to run batched, every image, then every clip, so each model loads once. Model load time is added where we recorded it; our text-generation runs don't carry a load measurement yet, so pipelines with a language-model stage are slightly optimistic. Nothing here is a single timed run of the whole pipeline, and we don't present it as one.