Prepare your inputs
Define the task, character, and creative constraints. Supply the conversation, images, audio, or storyboards it needs.
- Include relevant context
- Manage user consent and asset versions
- Keep the input structure consistent
DEVELOPERS
Build with
intelligence.
Make it yours.
Bring natural conversation, real-time characters, and video generation to your app.
API access available by request
Find your model.
Match the model to your task, inputs, and expected outputs.
Context-aware conversations that stay in character.
Multi-turn dialogueCharacter consistencyEmotion and social context
Role-play, AI companions, training scenarios, and synthetic persona research.
Input
Character profile, conversation history, and current message
Model output
Context-aware replies and action intents
Your app
Your app manages conversation history and session state, then sends the context needed for each request.
From input to experience.
Prepare the right context, call the model, and put the result to work.
Run the model
Choose a model for conversation, character video, or visual generation, then follow its API specification.
- Confirm supported capabilities
- Record inputs and generation settings
- Monitor output quality and response time
Deliver the experience
Your app handles display, playback, and follow-up actions. Use the result to guide the next interaction.
- Handle failures, retries and cancellation
- Maintain session and application state
- Log versions and user feedback
Start with your project.
Share your use case and sample inputs so we can discuss the right integration.
Tell us about your project
- What you’re building
- Your users, the problem you’re solving, and where the model fits.
- How you’ll use the model
- Your preferred model, source materials, output format, and expected usage.
- How you’ll judge results
- Sample tasks and your quality, latency, and cost requirements.
Request access
Tell us about your team, your app, and the model you need.
Confirm the details
Discuss model versions, capabilities, access, and operating requirements.
Build and test
Once approved, use the API documentation for your model to integrate and test.
Build on evidence.
Use research and task-based evaluations to test what works for your app.
Common questions
How do I get API access?
Select “Request access” and tell us which model you need and what you’re building. We’ll email you about available versions, access, and integration.
Who manages conversation history?
Your app stores conversation history and includes the relevant context in each request. OmniSoul responds using the character profile, history, and new message you provide.
When do we confirm API requirements?
During onboarding, we confirm the API format, authentication, capabilities, quotas, and deployment requirements for your model version.
How do I evaluate a model?
Test representative inputs against clear output requirements. For dialogue, check context and character consistency. For video, check appearance, motion, and audio. Measure quality, latency, and cost together.