KieAPI

Kie.ai is a unified API platform that lets you access OpenAI capabilities—GPT chat models for reasoning and coding, GPT Image models for image generation and editing, and Codex for code-focused automation—through a single API key. Instead of juggling separate providers, you can select the right model for each workload while keeping one integration surface.

For text and agent-style workflows, Kie.ai supports OpenAI Responses–compatible requests (POST https://api.kie.ai/codex/v1/responses) with streaming events and model IDs such as gpt-6-1-sol. You can tune response behavior with adjustable reasoning effort (low through max), add web search or function calling for agent tooling, and request structured output via JSON-schema in strict mode when your app needs predictable, parseable responses. The API also supports multimodal inputs, including text plus images and files in a single message.

For visual production, Kie.ai offers GPT Image models that support image generation and editing (including text-to-image and image-to-image workflows). For engineering automation, Codex models are positioned for tasks like repository-scale code assistance, code translation, syntax validation, and diagnostics.

Kie.ai uses pay-as-you-go billing, and pricing varies by model (example: GPT chat models are listed with token-based pricing; GPT Image models are listed “from” per-image pricing, with additional model-specific rates shown in the model library). Free trial credits are also mentioned for early workflow evaluation via new developer logs.

What can I build with the Kie.ai API?

You can build chat and reasoning workflows with GPT chat models, create and edit images with GPT Image models, and automate coding tasks with Codex models—all through one API key.

How do I call the API for GPT chat or Codex models?

Send a POST request to https://api.kie.ai/codex/v1/responses using the model ID from the model list (for example, gpt-6-1-sol). The endpoint follows the OpenAI Responses format and supports streaming events.

Can I control reasoning cost vs quality per request?

Yes. You can choose an adjustable reasoning effort level (from low to max) to balance speed, depth, and cost for each request.

Does the API support tools like web search or agent function calling?

Yes. You can enable built-in web search for up-to-date information or provide your own function tools for agent workflows; use one or the other per request.

What input types are supported?

You can send text, images, and files together in a single message, enabling multimodal workflows within one GPT-style request flow.

How does pricing work on Kie.ai?

Billing is pay-as-you-go and model pricing varies. Token-based pricing is shown for chat models, and GPT Image models are listed with per-image “from” pricing; free trial credits are also mentioned for new developer logs.