About
Ship your AI Agent. Understand the API calls behind it.
Tikway brings multi-model AI API access, bilingual integration guides, and cost tools together for developers building AI apps and agents.
Why it matters
An agent can make several model calls to plan, use tools, read results, and respond. Comparing the price of just the final answer can miss much of the workload. We help developers check the interface, model choice, and estimated usage before connecting a project.
What you can explore
• A public model directory to compare currently listed options and pricing information.
• An OpenAI-compatible API entry point: https://api.tikway.ai/v1
• English and Chinese setup guides, including SDK and chat-client configuration.
• A token cost calculator for estimating input/output costs and request volumes.
• Free MIT-licensed Python, Node.js streaming SSE, and offline cost-calculation examples.
Model identity is part of the promise
Our operating commitment is to provide the model indicated by the selected model ID, without disguising a different model as that choice. We also share a practical guide to checking model IDs, usage, and billing records.
Start with the resources
Models: https://tikway.ai/models
English guides: https://tikway.ai/en/guides
Cost calculator: https://tikway.ai/tools/token-cost
Open-source examples: https://github.com/lintt76/tikway-examples
The guides, calculator, and example code are free. Model API calls are paid according to the platform's current usage and billing rules. Available models, supported capabilities, and actual charges should be checked in the console.
We operate Tikway and welcome feedback on real integration problems: protocol setup, streaming responses, model selection, and agent task costs.