Chargement...
Loading...
Chargement...
Every creation plan (trailer, film) includes a production envelope: the real-cost budget for AI generations. Here is, action by action, what that envelope covers.
Our revenue comes from creation plans (you buy a result, not abstract credits) and viewing subscriptions. The commercial margin lives in the plan; the production envelope is deducted at real cost of the AI engines. On viewing, 50 % of the net is paid out to creators of the films watched.
AI analysis of your screenplay with feedback and suggestions
Generate one storyboard frame from scene description
High-quality film still generation
5-second AI video clip generation
15-second AI video clip generation
Complete trailer assembly with AI
AI music composition (30s-2min)
AI sound effect generation
Clone a voice for character dialogue
For token-based actions (text analysis, script generation, etc.)
| Model | Provider | Input / 1M tokens | Output / 1M tokens |
|---|---|---|---|
Claude Haiku 4.5 | anthropic | 0.08 cr | 0.40 cr |
Claude Sonnet 4.6 | anthropic | 0.30 cr | 1.50 cr |
Claude Opus 4.6 | anthropic | 1.50 cr | 7.50 cr |
GPT-4o | openai | 0.25 cr | 1.00 cr |
GPT-4o Mini | openai | 0.01 cr | 0.06 cr |
cr = credits. 1 credit = 1,000,000 µ-credits. Costs reflect real AI provider rates.
Each plan includes an internal real-cost budget (video generation, AI tokens, voice, QA). You work however you want — Auto, Guided or Pro — and each generation is deducted from this envelope at the real cost of the chosen engine. A gauge is always visible.
Each AI action has a credit cost based on the real provider cost (Anthropic, video engines, etc.). For language models, the cost depends on the number of tokens processed.
If an AI action fails for a technical reason (timeout, server error), the envelope is not charged. No deduction if the action produces no result.
The film is never blocked: you can finish in Auto mode (less hungry) or add a pack of extra minutes. You can never produce at a loss.