Retro anime sequence made with Scenario MCP: >Opus 5.5 blocks every shot in Blender >GPT Image 2.5 designs the pilot and the mech >Seedance 2.5 animates it, ElevenLabs scores it

Powered by @Scenario_gg We help creative teams orchestrate and scale everything they produce. Try it out today 👇

Looks familiar

What do you estimate 1 minute costs total? I've always appreciated your pricing transparency and I guarantee whatever it is it's 1/1,000 cost of traditional human workflows 🤣

Entire workflow costed around $15

Insane value, mind blowing.

Badass! Nice workflow. I'll have to give that a try.

How much skill is needed in prompting to get this level of result?

Zero. Just use the recipe here

Cuts are too sudden and the whole thing feels it's going too fast.

fer real tho you and @boneGPT have a chance for an epic cross over ngl

Such sick results always!

Remember when we used to prompt videos to existence 😂😂 so vintage

What would you estimate is the cost per minute of footage?

Refine your direction, add a good harness or skill, and Sonnet 5.5 on the Scenario MCP delivers excellent 1-2 minute videos for $15-25, Claude tokens and Scenario credits included.

Blocking every shot in Blender before generating is the step most AI shorts skip, and it's why the mech keeps its scale from cut to cut. That's real previs, with the models as crew. Did Seedance follow the Blender camera moves closely, or did you have to re-prompt the motion?

Very impressive! What’s your experience from using Blender steering with models other than Seedance 2/2.5? In my experience it’s very unreliable using anything else.

Wild collab, this actually goes hard

How are you running Seedance 2.5? Elevenlabs keeps blocking it for me

How did you get such good blocking with Opus? Was it all prompting or you gave it video reference animations as well?

Using Blender for shot blocking before the image and video models touch it is smart. That's probably why the camera work feels so consistent across shots.

dooop

The anime style is derivative and this is pretty much lifted nearly entirely from Pacific Rim and other mecha anime.

blocking in blender first is the smart move, the video model gets real camera moves instead of guessing. and each step being its own tool means you can swap seedance out later without redoing the rest. wrote this up here: