Free planning tool
Turn your scene into an editable production brief, check your reference order, and estimate the generation budget. This tool prepares prompts and budgets; it does not render video.
Everything here stays in your browser. Nothing is uploaded or sent to a video model.
Name the files you plan to use, their purpose, and (for video or audio) the clip length. No upload and no URLs are stored — this only orders them as @Image1, @Video1, @Audio1.
No references yet. With none, the planner targets the text-to-video endpoint.
Fill the brief and choose Build prompt to get an editable English production brief you can copy into a provider.
Fal Seedance 2.5 rate for 480p and 720p. Changes update the estimate immediately.
US-hosted is a hosting region, not a search market. Billing input video duration comes from the lengths entered in the reference plan.
Per attempt · 720p 16:9 · 5s billed
$2.31
Total for 1 attempt
$2.31
Tokens = width × height × 24 × (output + reference video seconds) ÷ 1024; cost = tokens ÷ 1000 × rate × reference multiplier × attempts. Model: bytedance/seedance-2.5/text-to-video (text-to-video). Audio on or off costs the same. This is a formula-based estimate, not a subscription price or a billing quote. Source checked 2026-10-04. The provider also shows an approximate per-second figure that does not match this formula; this tool uses the formula and the pixel dimensions.
These open the provider in a new tab. Rendering happens there and may require credits or a paid account; provider signup credits do not automatically include Seedance 2.5.
Ready-to-copy planning briefs. They are prompts, not generated videos.
Scene: An unbranded ceramic mug on a wooden desk, soft morning window light. 0–2s: Begin with a close shot of the mug handle; slowly move around the mug. 2–5s: Reveal the full mug and a small curl of steam; keep the mug shape and desk position unchanged. Camera: One smooth continuous move, no cuts. Audio: Gentle room ambience, no music or speech. Keep: No logos, captions, people, or hands.
Scene: An original adult gardener standing beside a small greenhouse. 0–3s: The gardener looks at a newly opened flower; medium shot, still camera. 3–7s: The gardener says: "This one opened overnight." Keep the voice calm and natural. 7–10s: Slowly move towards the flower while the gardener stays in frame. Audio: Spoken line plus quiet outdoor ambience, no background music. Keep: Consistent clothing and greenhouse layout; no captions.
Reference list: @Image1 = your own unbranded lamp photograph. @Video1 = your own camera-movement clip. Use the lamp design from @Image1 and the slow camera movement from @Video1. 0–4s: Show the lamp switched off on a plain table. 4–7s: The lamp turns on and casts a soft warm pool of light. 7–10s: End with the camera holding on the illuminated lamp. Keep: Exact lamp shape and tabletop position. No text, brand marks, or added objects. Audio: Quiet room ambience; no dialogue or music.
ImageToVideoPrompt is an independent planning tool. It is not ByteDance, Dreamina, or Fal, is not an official generator, and does not produce video. Your brief and reference labels stay in your browser.
Describe the scene, subject, action, camera, and audio, then list the references you plan to use as @Image1, @Video1, or @Audio1.
Read the generated English brief, fix any missing-subject or unbound-reference notes, and copy it into your chosen provider.
Pick resolution, aspect, length, and attempts to estimate the Fal API cost, with the video-reference multiplier shown explicitly.
Three things move the estimate: output resolution, the length of the shot, and whether you supply a video reference. Higher resolution and longer clips increase the token count, and a video reference bills both the input and output duration at a 0.6 multiplier — cheaper per second, but charged across the combined length. Additional attempts multiply the per-attempt estimate. Audio does not change the price. Always confirm the current rate at the official sources listed with the calculator, because provider pricing and reference limits can change.
Working on an image-to-video workflow? The AI image description generator turns a photo into an English scene description, and the image-to-video prompt library shows how to turn that description into motion. Use the AI model API cost calculator when you also need to budget the surrounding model calls.
No. This page prepares a prompt and estimates a budget. It does not call a model, does not upload your files, and produces no video. You build the prompt here, then render on the provider you choose.
For 480p and 720p, tokens = width × height × 24 × (output seconds + reference video seconds) ÷ 1024, and cost = tokens ÷ 1000 × rate × reference multiplier × attempts. The standard Fal rate is $0.0214 per 1000 tokens and the US-hosted rate is $0.02568, checked 2026-10-04. When a video reference is supplied the provider multiplies the price by 0.6 and bills both input and output video duration.
Fal’s reference-to-video pricing multiplies the price by 0.6 when video inputs are provided, so the per-second figure is lower — but it bills both the input video and the output video duration, and the token count grows with the combined length. The estimate shows the multiplier and the billed seconds so the total stays clear.
No. Per Fal, the cost of video generation is the same whether audio is generated or not, so this planner does not discount the estimate when audio is off.
No. ImageToVideoPrompt is an independent planning tool, not ByteDance, Dreamina, or Fal. The buttons labelled "Open Dreamina" and "Open Fal playground" open the provider’s own site in a new tab, where any credits or account requirements apply.
Neither. They are original prompt templates written for this tool. They are text you can copy and adapt, not rendered results and not third-party media.
Nowhere. The brief, the reference labels, and the estimate are all computed in your browser. Nothing is uploaded and no prompt or reference name is sent to a model or stored on a server.
