FABLE
Fable
Code and document analysis. Messages API connection and Claude Code setup.
Chat, Studio and API
AI models for code, images and video.
Use GPT and Claude in chat, create images and video in Studio, or connect your own client. Manage keys, balance and request logs in one CLODEx account.
Download CLODEX IDE →
API endpoint
Base URL
https://clodex.xyz/v1
Responses API
Streaming
Tools
Usage logs
POST /v1/responses { "model": "gpt-6-astra", "input": "Review this function" }
Coding · Video · Image
Live model catalog
Model-specific rates
Billing units are listed in the catalog
OpenAI Responses
Compatible endpoint
Usage tracking
Logs and balance in one account
New on CLODEX
GPT-6 Astra and Claude Fable. Choose for your task.
See capabilities, exact IDs and request examples. Codex and Claude Code have separate setup guides because they use different protocols.
Live now
FABLE
GPT-6 ASTRA
GPT-6 Astra
A reasoning model for multi-step code tasks. Text and image input, text output.
Models & pricing
Models and connection details.
Open a model card for its ID, protocols, billing unit and a practical guide.
Cache discounts GPT −20% Astra −30% GLM-5.3 −30% Conditions
Discounts apply automatically only to cached input tokens when the upstream reports a cache hit. Uncached input and output keep the regular price.
Supported GPT models: 20%. GPT-6 Astra and GLM-5.3: a separate 30% discount. Discounts do not stack.
Prices auto-update
View pricing
Coding models
$0.5/ 1M
Fixed Usage price
Cached input$0.35 / 1M · 30%
Details and setup
LLM routes for coding agents, reasoning, tools and everyday development.
Get an API key$0.25/ 1M
Fixed Usage price
Cached input
$0.20 / 1M · 20% off
$1.2/ 1M
Fixed Usage price
APIModel routing
Details and setup
LLM routes for coding agents, reasoning, tools and everyday development.
Get an API key$1/ 1M
Fixed Usage price
APIModel routing
Details and setup
LLM routes for coding agents, reasoning, tools and everyday development.
Get an API key$0.75/ 1M
Fixed Usage price
API
Anthropic + OpenAI
Details and setup
Claude for code and document analysis. See the card for the exact route and rate.
Get an API key
$0.09/ 1M
Fixed Usage price
API
OpenAI-compatible
Details and setup
Cost-efficient coding, long-context work and everyday agent automation.
Get an API key
$0.15/ 1M
Fixed Usage price
Cached input
$0.105 / 1M · 30% off
Details and setup
High-volume coding and agent workloads with a 30% cached-input discount.
Get an API key
$0.17/ 1M
Fixed Usage price
APIModel routing
Details and setup
LLM routes for coding agents, reasoning, tools and everyday development.
Get an API key
$0.08/ 1M
Fixed Usage price
Cached input$0.0704 / 1M · 12%
Details and setup
LLM routes for coding agents, reasoning, tools and everyday development.
Get an API key$0.07/ 1M
Fixed Usage price
Cached input
$0.056 / 1M · 20% off
Details and setup
Local changes, data transformations and tests with an explicit contract.
Get an API keyVideo models
Dynamic quoteper generation
Quote depends on duration and parameters
Video API
Async jobs
Details and setup
Text-to-video or image-to-video; choose duration and framing in Studio.
Get an API key
Dynamic quoteper generation
Quote depends on duration and parameters
Video API
Async jobs
Details and setup
Google Veo video: see task settings for supported modes, duration and resolution.
Get an API keyImage models
$0.03/ image
Quality-dependent image price
Quality modes
Auto · Low · Medium · High
Details and setup
Image generation and editing with auto, low, medium and high quality levels.
Get an API key$0.10/ request
Per-request price
API
OpenAI-compatible
Prices are synchronized with the current Fixed Usage catalog. Cached input is 20% cheaper for supported GPT models, 30% cheaper for GPT-6 Astra, and 30% cheaper for GLM-5.3; completed request cost is recorded in usage logs.
Service updates
What changed in CLODEX.
Latest CLODEX highlights with a complete public @clodex_api archive available as crawlable HTML, RSS and JSON.
Видеомодели и Claude Antigravity работают стабильно
Open update →🥳 Обновили Claude MAX и починили генерацию видео
🌟 Больше моделей в Claude MAX Добавили оригинальные модели: • claude-sonnet-5 • claude-sonnet-4-6 • claude-haiku-4-5 (claude-haiku-4-5-20251001) • claude-fable-5.1 Все они доступны напрямую через ключи CLAUDE_MAX и в Claude Code. 😀 Исправили про…
Open update →🍟 CLODEx — большое обновление! Новые GPT-6 Ultrafast, Fast-модели и улучшения Codex
С 15 по 20 октября мы добавили новые ИИ-модели, ускорили работу Codex и обновили инструменты генерации контента. Рассказываем о главном 👇 🧠 Новые GPT-6 Ultrafast и Fast Теперь в CLODEx доступны новые скоростные модели: 🧠 Ultrafast — для максима…
Open update →Gemini-4-argon стабилизирован и снова в бою
Open update →🪙 Стоимость моделей Muse была снижена
Актуальные цены на модели Meta в Pricing
Open update →Quick start
From account to first request in four steps.
Create a key, choose an available model, configure your client and check one request before working on a project.
-
Create an API key
Register and issue a key in the CLODEX dashboard.
Create an account -
Top up balance
Check your model’s rate and the payment options in your account.
Open wallet -
Connect your client
Codex and OpenAI clients use the /v1 base; Claude Code uses the root domain.
Connection guide -
Start building
Send a short request, then check its result and charge in the log.
Choose a model
Integrations
One endpoint — all your tools.
Use one CLODEX key across desktop clients, coding agents and SDKs.
CLODEX IDE
CLODEx desktop client. Installing the client and paying for model inference are separate.
Open setup guide →
Codex App
Install configuration for macOS, Windows, Linux and WSL.
Open setup guide →
Claude Code
Root domain without /v1, a CLODEx key and an exact Messages model ID.
Open setup guide →
Cursor
Use the CLODEX base URL and key in your editor workflow.
Open setup guide →
OpenCode
Configure an OpenAI-compatible provider for agent tasks.
Open setup guide →
Cline
Bring CLODEX models into your existing VS Code workflow.
Open setup guide →
OpenAI SDK
Configure the base URL, key and model, then check the SDK methods you need.
Open setup guide →
Anthropic SDK
Use supported Anthropic-compatible routes from one account.
Open setup guide →Connection hub
Choose your client.Find its settings.
Each guide gives the base URL, authentication, a request example and response handling for the relevant protocol.
LLM GATEWAY
BASE URLOpenAI-compatible
https://clodex.xyz/v1
Unified LLM API
Choose Chat Completions, Responses or Messages.
Connect a model
Build with your tools
Base URL depends on the client protocol. Each guide has the exact settings.
All API documentation ↗CLODEX IDE
A free IDE already connected to CLODEX API.
The fastest way to start: install the client, sign in and select a model. Manual providers remain available when you need them.
- ✓Automatic endpoint and model catalog setup
- ✓Balance and usage visible inside the interface
- ✓Agent workspace, terminal and file tools
- ✓OpenAI, Anthropic, Ollama and custom endpoints remain available
clodex-app / workspace
$12.48 · GPT-5.6 Sol
agent.tsxapi.ts
14 const client = new OpenAI({
15+ baseURL: "https://clodex.xyz/v1",
16+ apiKey: process.env.CLODEX_API_KEY
17 })
19+ const response = await client.responses.create({
20+ model: "gpt-5.6-sol",
21+ input: task
22 })
➜ clodex-app bun run check
✓ TypeScript check completed
✓ 6 tests passed
✓ TypeScript check completed
✓ 6 tests passed
API first
A request you can verify.
A Responses API example. Use the client-specific guides above for other protocols.
curl · Responses API
gpt-5.6-sol
curl https://clodex.xyz/v1/responses \
-H "Authorization: Bearer $CLODEX_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-5.6-sol",
"input": "Review this function"
}'
Control & reliability
Inspect requests and spending in one place.
Use the log to match a model, request time and charge. Retain a safe request ID for troubleshooting.
Operational help
Get help diagnosing endpoint, account and payment issues.
Usage logs
Track requests, models, tokens and cost in one place.
Clear billing
Balance, top-ups and spending remain visible.
Keys and limits
Separate keys and control access for different clients.
API documentation
Connection guides for clients, SDKs and operating systems.
Support
Get help with connection, billing and account issues.
CLODEX IDE is open source. Inspect the client, contribute improvements or configure another provider manually when needed.
View on GitHubFAQ
Questions about CLODEX LLM API.
How access, keys, billing and client connections work.
What is CLODEX?
CLODEX is a unified LLM API gateway with GPT, Claude, Gemini, Grok and other models, one balance and usage logs.
What base URL should I use?
OpenAI-compatible clients use https://clodex.xyz/v1. Anthropic SDK and Claude Code use https://clodex.xyz without /v1.
Can I use one key for different models?
Yes, when the models are available to the key group. For production, create separate keys for each application or environment.
Does CLODEX support streaming and tools?
Support depends on the model and route. Streaming needs a terminal-event check; tools need argument and authorization checks on your backend. One successful text answer does not verify those features.
Where can I see current prices?
Use the public pricing page. It shows input, output, cache and Fixed Usage billing where applicable.
How do I contact support?
Use Telegram @clodex_labs for account and API support. Product news is published in @clodex_api.
Does one account replace every model subscription?
No. CLODEX provides access within its own service and API. A subscription on a model owner’s website is a different product. Check the selected model, billing unit and available access before paying.
What should I check before using CLODEX from my region?
Check site access from your network and the payment options available in your account. Confirm model access with your own key before integrating. A listing does not guarantee every network or bank card is supported.
Connect your first client.
Choose the guide for your application, then create a dedicated key for it.