🚀 GLM-5.2 + Doubao Seed 2.1 are live now

One API.
Every model.
From ¥1.84 / 1M.

OpenAI-compatible gateway to GLM-5.2, Doubao Seed 2.1 and GLM-4-Flash. Drop-in replacement for OpenAI/Claude — just change the base URL.

Get free API key → Read docs
cURL
Python
Node.js
# Drop-in replacement for OpenAI/Claude curl https://onefreeapi.cloud/v1/chat/completions \ -H "Authorization: Bearer sk-ofa-..." \ -H "Content-Type: application/json" \ -d '{ "model": "glm-4-flash", "messages": [{"role":"user","content":"Hello!"}] }'
1
Models available
¥1.84
Starting price / 1M input tokens
5
Global edge nodes
1,000
Free tokens on signup

Models

Top open & proprietary models, all behind one API. Switch models by changing one parameter.

Kimi K3 NEW
Moonshot's flagship. 256K context. Coding & reasoning.
$0.003/1K in · $0.006/1K out256K ctx
GLM-5.2
Zhipu AI · 1M context. Long-document analysis.
$0.003/1K in · $0.006/1K out1M ctx
DeepSeek V3
Open-source flagship. Best price/performance.
$0.001/1K in · $0.002/1K out128K ctx
Doubao Seed 1.6
ByteDance. Fast, cheap, multilingual.
$0.0015/1K in · $0.003/1K out256K ctx
Qwen 2.5 FREE
Local inference. Test & prototype free.
$0 · 100 calls/day32K ctx
More coming
GPT-4.1, Claude Sonnet 4.5, Gemini 3 — all in Q3.
Stay tunedRoadmap

Simple Pricing

Pay-as-you-go tokens. No monthly commitment. Free tier forever.

Free
$0/mo
Test & prototype. No credit card.
  • 1,000 free credits on signup
  • Works with OpenAI SDKs, zero code change
  • 1 API key
  • Community support
Start free →
Pro
$14/mo
For production apps. All models.
  • 14,000 credits included
  • All premium models (Kimi K3, GLM-5.2)
  • 5 API keys
  • Usage dashboard
  • Email support
Get Pro →
Enterprise
$139/mo
Scale with SLA & support.
  • 139,000 credits included
  • Unlimited API keys
  • Multi-region routing
  • Priority support (SLA)
  • Custom invoicing (USDT/WeChat/Alipay)
Contact us

Drop-in Replacement

If your app uses OpenAI SDK or Anthropic SDK, you're 30 seconds away.

Python (OpenAI SDK)
# pip install openai from openai import OpenAI client = OpenAI( base_url="https://onefreeapi.cloud/v1", api_key="sk-ofa-...", # your OneFreeAPI key ) response = client.chat.completions.create( model="glm-4-flash", # more models as they go live messages=[{"role": "user", "content": "Hello!"}], ) print(response.choices[0].message.content)

Tools & Agents

Point your coding tool or AI agent at OneFreeAPI — one key, every model.

WorkBuddy / AI Agents

In WorkBuddy and similar agents, pick Custom (OpenAI-compatible API) in “Add model” and fill in these four fields. The model name must be a bare id — writing openai/glm-4-flash returns Model not found. Heads-up: some clients’ “Test connection” only calls GET /v1/models, which is public here, so a wrong key can still show green — send one real message to confirm.

add model
供应商 自定义（OpenAI 兼容协议 API） 接口地址 https://onefreeapi.cloud/v1/chat/completions API Key sk-ofa-... # create in dashboard, shown only once 模型名称 glm-4-flash # bare id only, not openai/glm-4-flash

OpenClaw

Native support for OpenAI-compatible endpoints. Edit ~/.openclaw/openclaw.json:

openclaw.json
{ "models": { "default": "onefreeapi/glm-4-flash", "providers": { "onefreeapi": { "baseUrl": "https://onefreeapi.cloud/v1", "apiKey": "sk-ofa-...", "api": "openai-completions", "models": [ { "id": "glm-4-flash", "name": "GLM-4-Flash", "contextWindow": 131072, "maxTokens": 8192 } ] } } } }

Codex CLI

Edit ~/.codex/config.toml and export ONEFREEAPI_API_KEY=sk-ofa-.... Note that wire_api = "chat" must be set explicitly (we serve Chat Completions; some newer Codex builds only support the Responses API).

config.toml
model = "glm-4-flash" model_provider = "onefreeapi" [model_providers.onefreeapi] name = "OneFreeAPI" base_url = "https://onefreeapi.cloud/v1" env_key = "ONEFREEAPI_API_KEY" wire_api = "chat"

Claude Code

Claude Code speaks the Anthropic Messages protocol. Bridge it to OneFreeAPI via claude-code-router: npm i -g @musistudio/claude-code-router, then edit ~/.claude-code-router/config.json and launch with ccr code.

config.json
{ "Providers": [ { "name": "onefreeapi", "api_base_url": "https://onefreeapi.cloud/v1/chat/completions", "api_key": "sk-ofa-...", "models": ["glm-4-flash"] } ], "Router": { "default": "onefreeapi,glm-4-flash" } }

Any OpenAI-compatible tool

Cursor, Cline, Continue, Cherry Studio, ChatBox, Dify, n8n — any tool or agent that accepts a custom OpenAI Base URL. Fill in these three fields:

settings
Base URL: https://onefreeapi.cloud/v1 API Key: sk-ofa-... # create in dashboard, shown only once Model: glm-4-flash # check GET /v1/models for the current list

Frequently Asked Questions

Everything you need to know about the OneFreeAPI gateway.

What is OneFreeAPI?
OneFreeAPI is an OpenAI-compatible LLM API gateway. One API key gives you access to 10 models across Zhipu, ByteDance, DeepSeek and Moonshot — GLM-5.2, GLM-5.3-Flash, GLM-4-Flash, Doubao Seed 2.1 Turbo/Lite, DeepSeek V4/V4.1 and Kimi K3.1-Pro — with unified billing and automatic failover between upstream providers.
How do I migrate from OpenAI?
It takes about 30 seconds. Keep using the official OpenAI SDK (Python, Node.js) or plain HTTP, set base_url to https://onefreeapi.cloud/v1, and replace the API key with your OneFreeAPI key (sk-ofa-...). No other code changes are needed — chat completions, streaming and model listing follow the OpenAI schema.
Which models can I use?
Current lineup includes Kimi K3 (Moonshot, 256K context), GLM-5.2 (Zhipu, 1M context), DeepSeek V3, Doubao Seed 1.6 (ByteDance) and GLM-4-Flash (free tier). The live list is always available at GET /v1/models.
Is there a free tier?
Yes. Every new account gets free credit after verifying your email — no credit card required — plus 100 free calls per day on free-tier models. Paid plans start at $14/month for 14,000 credits.
What payment methods do you accept?
Pay-as-you-go token top-ups and subscriptions can be paid with USDT, WeChat Pay or Alipay. Enterprise plans support custom invoicing — contact support@onefreeapi.cloud.