MODELS & ROUTING
Different models. A common language.
Organize models and routing policies around each use case, giving every application the capabilities it needs.
- Shared interface conventions
- Routing by scenario
- Model scope agreed in your plan
THE ENTERPRISE AI GATEWAY
Intelligence.Part of everyday life.
Bring models, compute and applications together. One gateway for access and orchestration, so your team can focus on what comes next.
Unified API · Domestic models · Flexible deployment
One city. Every layer connected.
01
One entry pointConnect models and products
02
Shared controlUnderstand access and usage
03
Your deploymentDefine resources and boundaries
Start with the right model
Example model families · Confirm available versions and deployment scope before integration- DeepSeek
- Qwen
- GLM
- Kimi
- MiniMax
02 / THE GATEWAY
The complexity is ours. The possibilities, yours.
Follow a request to see how t.city connects your application to a model. Bring different workflows to one entry point, with routing, access and usage organized around your needs.
01Your application
Turn this product guide into a set of customer FAQs.
t.city
Unified gateway
- 01Access checks
- 02Scenario routing
- 03Usage records
Retrieval · Content generation
03Model response
◈Qwen
Organized into three themes: features, integration and usage. Each answer can link back to the original source for review.
Usage records
Interactive illustration · Preset content, no live model calls
03 / BUILT FOR YOUR TEAM
From one application to your whole team.
Access is only the beginning. Give every team a common foundation for the work around their models.
TEAMS & GOVERNANCE
Every request. Clearly accounted for.
Manage access, quotas and usage by team and project, with a shared view for ongoing operations.
- Keys and member access
- Project quota allocation
- Usage and cost reconciliation
RESOURCES & DEPLOYMENT
Your boundaries. Your deployment.
Plan API access or a private deployment around data location, resource isolation and operational requirements.
- Deployment region assessment
- Dedicated resource planning
- Logging and operations agreements
04 / DEPLOYMENT
One city, different ways to land.
Start with one API, then choose a deployment path around data boundaries, traffic and team governance. We confirm the details before delivery.
Your appProducts · agents · workflows
Service delivery scope
t.city gatewayRouting · quotas · logs
Model computeModels on demand
- Resource boundary
- Shared service boundary
- Suitable for
- Prototypes, light production, model experiments
- Confirm before launch
- Confirm quotas, region and billing
Agree on data location, model licenses, log retention and access permissions before delivery. Capabilities and responsibilities follow the agreed solution and service contract.
Discuss this pathCompany entity and regional operators
- Group entity
- 中国词元集团有限公司 (China Token Group Co., Ltd.)
- Mainland China
- 唐马供销有限公司
- Overseas
- China Token Cloud Computing Co., Ltd.
05START BUILDING
Start with your use case.
An application, a deployment, or a compute partnership. Tell us what you want to build, and we can work out a practical path together.
Share your requirements
Outline your use case, model needs, deployment region, and existing systems.
Define the plan
Confirm available models, deployment options, service scope, and pricing together.
Integrate and launch
Once the plan is agreed, coordinate integration, testing, and launch.
Prefer a direct conversation?ceo@ku0.com
PROJECT BRIEF
Make the first conversation count.
Describe what you need and prepare an email to start the conversation.
GOOD TO KNOW
Before you start.
How do I get access?
Email us with your use case, operating region, and deployment needs. Integration and testing follow after the service scope and plan are agreed. This site introduces the product and helps you start that conversation.
Which models are available?
Model selection depends on deployment region, model licensing, and the service plan. Mainland-China workloads prioritize domestic open-source and open-weight models. The agreed plan determines actual availability.
Can I deploy in my own environment?
We can discuss private deployment and local inference. Model suitability, hardware, networking, and data boundaries require joint assessment. Delivery and support are defined in the agreed plan.