codeflare
Get started
YOUR MODELS. ONE CONNECTION.

One connection,
connect your
AI workflow.

AI API Gateway
From your first line of code to your next creation.

Connect multiple AI models through one API, pay for actual usage, and manage project keys, model permissions and spending limits.

Pay as you goClear usageProtocol compatibility
Codeflare console screenshot in Chinese: balance, model usage and request logsProduct preview · Demo data

Your familiar models, in one workspace.

OpenAIAnthropicGoogleDeepSeek
See the catalog for currently available models

Less integration work,
more room to create.

Model access, keys and usage,
together in one clear workspace.

Different models. One starting point.

Choose models for text, coding and reasoning, with protocol-specific configurations.

OpenAIAnthropicGoogleDeepSeek
Explore the catalog

Every app, under your control.

Create project-specific keys with model permissions, spending limits and expiration dates.

Manage API keys

Clear usage. Confident decisions.

Explore tokens, latency and cost breakdowns, from overall trends to individual requests.

View usage

Connect your tools,
keep your familiar workflow.

SDKs, Claude Code and Codex CLI,
with clear, protocol-specific setup guides.

View integration guide

Frequently asked questions

Is this a working API gateway?

Once an available channel is configured, use your API key to call models. Verify your email and add funds first.

How does billing work?

Usage is charged against your prepaid USD balance. Each request uses the price at admission. See the catalog for rates.

Which integrations are supported?

Configuration examples are available for cURL, Python, JavaScript, Claude Code and Codex CLI. Models are filtered by their supported protocols.

Do you store prompts or responses?

We retain request metadata for 90 days and do not record prompts or responses. Upstream storage follows request options and provider policies.

Your next idea starts here.

Create an account and connect to your models.

Get started