About TokenRa:
Connect Chinese AI models to your product.

What is TokenRa?

TokenRa is a unified AI API gateway that helps developers access text, image, video, and audio models through one account and API key. Channel selection, model management, usage, and billing are organized in one console.

It is designed for developers and teams that need to connect multiple models quickly, compare costs, or build a unified model layer for a product.

How do you integrate models?

Start in the Model squareto review available models and IDs，then readDeveloper documentationto complete API authentication and integration。Compatible text models use a unified OpenAI-compatible interface；image and video models may use their own asynchronous or media endpoints。

The typical flow is to register, create an API key, confirm model and channel availability, send a test request, then configure timeouts, retries, budgets, and logging for production.

Which models are supported?

The public model catalog includes DeepSeek, Kimi, GLM, Omen Alpha, Ox Alpha, Seedance, and GPT Image models and model guides. Exact versions, model IDs, input modalities, and available channels are shown in the Model catalog and live console.

For example, Ox Alphais suited to coding and agentic work，Omen Alphaprovides coding and reasoning API access；Pricing and capabilities differ by model.

Billing, pricing, and refunds

TokenRa uses usage-based billing and does not charge a fixed platform subscription fee.Billing units for input, output, cache reads, image tasks, and video tasks can vary by model and channel.Public pages provide reference rates; checkout and the console show the actual price.

Top-ups, balances, promotions, and refunds are governed by the current terms and account transaction records.For billing questions, contact support with account details, request time, and request ID so the team can review verifiable usage records.

Data and privacy

Keep API keys server-side; never place them in browser code, public repositories, or client applications. Prompts, images, videos, and outputs sent to upstream models may be processed under the applicable provider policy; before sending sensitive or regulated information, readPrivacy Policy and applicable provider terms.

TokenRa does not develop or own third-party models. Model names and trademarks belong to their respective owners.

Support and contact

For account, API key, billing, or API issues, email support@tokenra.io，with the relevant request context. You may also submit a ticket through the console when enabled.

Frequently asked questions

How does TokenRa differ from calling an official API directly? TokenRa provides one account, channel management, and billing view, while upstream providers determine model capabilities, limits, data policies, and availability.

Can pricing change? Yes. Model versions, channels, promotions, and instance configuration can affect pricing; use the live console as the source of truth.

How do I start? Visitthe registration page, create an API key, then use the documentation and Model square to choose a suitable model.