How do you handle interface differences, configure channels, and understand usage and costs across models? BitCloud API brings model access, access control, request routing and usage management together to connect business applications with model services.
API keys & access controlProtocol adaptation & request routing
Business applications
Internal systems
Development & testing
BitCloud API
Model access & usage management
Text model channels
Video model channels
Models & channels
Usage & costs
Call records
Access and management overview
Applications call models through configured interfaces and channels. Capabilities and parameters depend on the selected model and channel.
From model configuration to application calls.
Manage the models, channels, access rules and processes needed for application calls.
Manage models and channels together
Review model capabilities, input and output types, and pricing. Configure model-to-channel relationships and maintain the service catalog available to applications.
Model catalog
Channel configuration
Available model services
Adapt interfaces and route requests
Adapt request and response formats across model services. Handle routing, streaming and abnormal calls according to channel configuration, with records to support investigation.
Application request
Protocol adaptation & routing
Model channels
Manage keys and usage permissions
Manage API keys, users and groups, define access scope for different callers, and review and maintain settings through the console.
Callers
API keys
Access scope
A unified entry point preserves model and channel differences. Verify supported interfaces and parameters for each integration.
Call each model service in the way it requires.
Text responses and video generation follow different processes. Check model capabilities, parameters and response formats when integrating.
Text calls
Send requests according to the model and protocol configuration, receive standard or streaming responses, and handle call status and results.
Send request
Receive response
Video tasks
Create generation tasks, query their status and retrieve results. Configure duration, resolution and media inputs according to model requirements.
Create task
Query status
Retrieve result
Understand call records and the basis for costs.
Connect usage, billing and call records to help teams reconcile activity, investigate issues and review operations.
Usage records
Review model calls and associated usage. Filter records to understand how services are being used.
Billing & settlement
Handle billing and quotas according to service configuration. Video tasks support precharges, result settlement and failure refund workflows.
Logs & exports
Query call information including models, channels, costs and errors. Export filtered records for reconciliation and investigation.
Billing units vary by service and may include tokens or video duration. Prices, settlement and refunds follow the applicable rules.
Support your team’s integration and operations work.
Configure models, channels and management processes for real tasks, from application integration to service operations.
Connect models to multiple applications
Centralize model access and call credentials, provide applications with a unified entry point, and reduce scattered configuration maintenance.
Build an enterprise model service entry point
Deploy the gateway and management console in an enterprise environment to organize model channels, permissions and daily management.
Manage service usage and costs
Use call records, usage and costs for daily reconciliation, issue investigation and service operations.
Manage integrations or use model services directly.
Choose the entry point for your task: explore the software or use existing model services.
BitCloud API is software for model access and usage management. Token Platform is an entry point for existing model services. Open Token Platform to use existing services, or contact us about enterprise integration, deployment or management.
Are all model interfaces and parameters identical after integration?
No. The product provides protocol adaptation and unified management, but capabilities, context, parameters and response formats can still differ. Verify the scope supported by the model, channel and product version before integrating.
Which text protocols are supported?
Existing adaptations cover OpenAI Chat, OpenAI Responses, Claude Messages and Gemini text protocols. Available interfaces, streaming events and parameters depend on the version and configured channels.
Do video generation and text calls use the same process?
Not exactly. Video usually involves task creation, status queries and result retrieval. Check duration, resolution and media input requirements; supported parameters differ across models and channels.
Is every service billed by tokens?
No. Text and video may use different billing dimensions, such as tokens or video duration. Video tasks involve precharges, result settlement and failure refund workflows. Charges and refunds follow the applicable rules.
Will a channel failure trigger a seamless automatic switch?
The platform provides retry, channel switching and route tracing capabilities, but switching depends on configuration, alternative channels and request compatibility. Seamless recovery cannot be guaranteed for every failure. Validate the integration against business requirements.
Can it be deployed in an enterprise environment?
Self-hosting and multi-instance deployment capabilities are available. The plan depends on enterprise networking, databases, caching, storage, upstream channels and workload. Capacity and operational requirements must be established through the actual solution and validation.
How do I get started?
Open Token Platform to use existing model services. For an integration discussion, describe your application, target models or channels, calling patterns, expected load and deployment environment.
