Runway Gen-4.5. Cinematic shots that obey the prompt.
Runway's latest generation model is tuned for scene craft: strong motion, high visual fidelity, and prompt adherence that keeps the shot you asked for.
- 2 to 10 second clips
- 16:9 and 9:16
- First frame control
Runway built its reputation with filmmakers, and Gen-4.5 is that sensibility in model form: cinematic scene creation with convincing motion and close prompt adherence, from text or from a first-frame image. Clips run 2 to 10 seconds in landscape or portrait 720p, and seed support means a shot you like can be regenerated reproducibly while you vary one thing at a time.
Gen-4.5 generates picture only, and on Treza that is not a limitation: wire it into a pipeline with a text to speech node for narration and a music node for the bed, and the output is a finished piece. It bills 12 cents per second of output from prepaid credits, charged only on successful runs.
Cinematic scene craft
Framing, light, and motion that read like coverage from a set. The Runway aesthetic, on the same canvas as every other model.
Prompt adherence
The shot you described is the shot you get, which matters when a storyboard, client, or pipeline is upstream of the render.
Reproducible seeds
Fix the seed and iterate deliberately: same bones, one variable changed per run, instead of gambling the whole shot on each roll.
Short beats to full shots
2 to 10 second range, and the 2-second floor is useful on its own for inserts, transitions, and B-roll stabs most models cannot do economically.
What to make with Runway Gen-4.5
Real prompts, and what comes back. Every one runs in Treza chat or as a node on the canvas.
“A detective steps out of a taxi into rain, neon reflections, handheld feel”
A moody establishing shot with committed camera work.
“Slow dolly across a dinner table after the party, morning light”
A quiet cinematic beat with strong art direction.
“A 2-second whip pan transition between city and desert”
A short insert shot priced at a fraction of a full clip.
“First frame: product still. The camera pulls back to reveal the workshop it was made in”
Image-to-video from an exact opening frame.
“A lighthouse in a storm, waves crashing, 9:16 for Shorts”
Vertical cinematic footage for short-form.
“Same seed, same prompt, but golden hour instead of storm”
A controlled variation on a locked composition.
Runway Gen-4.5 is one node on a canvas
Chain it with language, image, and video models, then publish the whole pipeline as an API.
- Step 01
Draft the brief
A language model node expands a one-line idea into a detailed, on-brand prompt before it ever reaches the model.
- Step 02
Generate with Runway Gen-4.5
Run variants in parallel, compare against other models on the same prompt, and keep every result in run history.
- Step 03
Publish it as an API
The pipeline becomes a versioned endpoint. Your product calls it with the typed /invoke API or any OpenAI SDK.
Runway Gen-4.5, answered
What is Runway Gen-4.5 best at?
Cinematic shots that need to match an intent: storyboards, branded film beats, and B-roll where framing and motion quality carry the piece. Seed support makes it the most iterable model in the lineup for deliberate shot development.
Does Gen-4.5 generate audio?
No, it renders picture only. On Treza the usual pattern is a pipeline: Gen-4.5 for the footage, a text to speech node for narration, a music node for the bed, and the Sequence node to mix them into one file.
How does it compare to Veo 3.1 and Sora 2 Pro?
Veo 3.1 brings native audio and Google's polish, Sora 2 Pro brings 20-second clips with dialogue, Gen-4.5 brings Runway's shot craft, prompt adherence, and reproducible seeds. They share one picker on Treza, so run the same brief on two and keep the better take.
How much does Runway Gen-4.5 cost on Treza?
Treza uses prepaid credits with no subscription. Each clip is charged at Runway Gen-4.5's rate from your balance, and only successful generations are charged. Credit packs start at $5 and never expire.
Can I call Runway Gen-4.5 through an API?
Yes. Build a pipeline that uses Runway Gen-4.5 and publish it as a versioned HTTP endpoint. Call the typed /invoke API from your product, or point any OpenAI SDK at the OpenAI-compatible endpoint. If a better model ships later, swap it on the node without changing the API your product calls.
Your next prompt could be production.
Generate your first video, image, or draft today. Prepaid credits, no subscription.