Start Frame
Use an image to establish the opening identity, product, styling, environment and composition for image-to-video production.
Ruwana Video · Platform API live
AI video production with continuity controls built into the workflow.
Ruwana Video combines text-to-video and image-to-video production with explicit visual continuity tools. Start and End frames define the shot boundaries, Bind keeps selected identities and visual elements present across the production, and Multi-Shot structures longer ideas into controlled scene segments. The same production system is available through the live Ruwana Platform Video API.
Image-to-videoText-to-videoStart / End framesBindMulti-ShotPlatform API live
Build the video around explicit production inputs
Ruwana Video exposes the decisions that matter for continuity and art direction instead of hiding everything behind one text box.
End Frame
Add an optional final frame when the production needs to move toward a specific closing image rather than an unconstrained ending.
Text-to-Video
Create video from production direction without supplied frame references when the scene should be generated from the written brief.
Bind for consistency across the shot
Bind lets a Video production attach selected Ruwana Virtual Models or custom visual elements to an image-to-video job. The goal is identity and object continuity across the generated motion and across Multi-Shot segments.
Virtual Models
Bind an approved Ruwana model to the Video production when the same model identity needs to remain part of the scene.
Custom elements
Add an object or visual reference for the current production when a product or element must remain connected to the shot.
Multi-Shot production
Break a video into controlled scene slots instead of compressing a sequence into one ambiguous instruction.
Define the sequence
Create multiple shot segments within the total video duration.
Direct each shot
Give each segment its own action or scene direction while maintaining the overall production concept.
Keep bound elements
Selected Bind identities and objects remain part of the production contract across the shot sequence.
Camera & motion direction
Describe camera movement, shot type, subject motion, mood and scene behavior. Ruwana Video supports explicit camera-direction workflows rather than treating camera behavior as an afterthought.
Audio & delivery options
Video production can be configured with or without generated/native audio depending on the workflow. Platform delivery is 1080p or 4K. I2V native audio is optional; T2V includes audio capability in its base rate.
Live Platform API contract
Submit one idempotent Video request, then follow the returned durable request until it succeeds or reaches a terminal state. Provider implementation details remain internal to Ruwana.
POST /v1/video/generate
GET /v1/requests/{requestId}
Modes: i2v | t2v
Resolution: 1080p | 4K
Duration: 3–15 seconds
Retention: 7 daysPublic Video pricing
Billing is success-only and uses the effective video duration. Idempotent replay does not create a second debit.
| Mode | 1080p | 4K | Audio |
|---|---|---|---|
| Image-to-video | $0.22/sec | $0.30/sec | Optional native audio +$0.08/sec |
| Text-to-video | $0.28/sec | $0.34/sec | Included |
Request rules
Image-to-video
Multipart Start Frame is required. End Frame is optional. Bind supports up to three total bound elements. Multi-Shot supports 2–6 ordered shots within the 3–15 second effective duration.
Text-to-video
No Start or End Frame is accepted. Aspect can be 16:9, 9:16 or 1:1. Audio capability is included in the base T2V rate.
Ruwana Video is live on Platform.
Use the same organization API key, wallet, idempotent request model and seven-day media delivery contract as the other Ruwana production APIs.