LLM Gateway · Compatible with OpenAI and Anthropic

Modellix LLM

One gateway to leading language models

Use one Modellix API key for Chat Completions, Responses, and Messages. Requests are synchronous with optional SSE. Image and video generation continue to use the media API.

https://llm.modellix.ai

This is the text gateway host. OpenAI-compatible clients should use https://llm.modellix.ai/v1.

Get an API keyView LLM pricingRead the docs

Quickstart

Connect your client in three steps

Set the base URL, provide a Modellix API key, and set model to provider/name.

  1. 01 · Key

    Create an API key

    Create it under Console → API Key, then send it as Bearer or x-api-key.

    Console → API Key
    Authorization: Bearer mdlx-xxxxxxxx
  2. 02 · Base URL

    Choose the host for your protocol

    OpenAI-compatible clients include /v1; Anthropic and Claude Code do not.

    https://llm.modellix.ai/v1
  3. 03 · Model

    Pass a model ID

    Always use provider/name. Refer to the product documentation for available models.

    openai/gpt-5.6-sol

Gateway capabilities

A gateway built for common clients and agents

One key covers Chat Completions, Responses, and Anthropic Messages.

  • Three mainstream protocols

    Chat Completions, Responses, and Messages each have their own URL and request body. Choose the protocol for your client and do not mix their fields.

  • Compatible with common clients

    OpenAI and Anthropic SDKs plus Codex, Claude Code, Cursor, OpenCode, and more connect by overriding the base URL. No Modellix-specific SDK is required.

  • One Modellix key

    Bearer and x-api-key are both supported. Use a key from Modellix Console, not a provider-issued key.

  • Synchronous with optional streaming

    Text calls return synchronously. Setting stream: true enables the protocol-specific SSE stream. Media generation remains asynchronous.

  • Consistent model IDs

    Set model to provider/name, such as openai/gpt-5.6-sol or anthropic/claude-sonnet-5.

  • Transparent per-token pricing

    Successful responses are billed from usage. Models with Input Context tiers show every tier in the price table.

Protocol selection

Choose the protocol before sending a request

Changing only the URL is not enough. Chat Completions uses messages, Responses uses input, and Messages requires max_tokens.

POST /v1/chat/completions · For the OpenAI SDK, Codex, Cursor, and most chat clients. The base URL must include /v1.

curl -sS "https://llm.modellix.ai/v1/chat/completions" \
  -H "Authorization: Bearer ${API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-5.6-sol",
    "stream": false,
    "max_tokens": 256,
    "messages": [{"role": "user", "content": "Introduce yourself in one sentence"}]
  }'
Chat CompletionsResponsesMessages
Path/v1/chat/completions/v1/responses/v1/messages
Primary inputmessagesinputmessages + optional system
Length fieldmax_tokens / max_completion_tokensmax_output_tokensmax_tokens (required)
Base URLhttps://llm.modellix.ai/v1https://llm.modellix.ai/v1https://llm.modellix.ai

Model catalog

Models and pricing

Prices are in USD per 1M tokens. The Context column shows the model’s supported window. Some models use Input Context pricing tiers; the number of tiers varies by model and is independent of the context-window size.

Cost calculator

Estimate one call

The input-token count automatically selects the matching Input Context tier.

openai/gpt-5.6-sol
Input tokens
Output tokens
Cached Read tokens
Cached Write tokens
Input Context ≤ 272KPriced by this request’s Input Context
Input$0.288
Output$0.036
Cached Read$0
Cached Write$0
Estimated Modellix cost$0.324

Original price: approximately $0.36 , approximately 10% less

Price table

Filter by provider
Input modality
ModelModalitiesContextInput ContextInputOutputCached ReadCached WriteDiscount
Claude Fable 5.1
anthropic/claude-fable-5.1
TextImage
Text
1M—$10.00$9.00$50.00$45.00$0.25$0.225$15.25$13.7210% OFF
Claude Haiku 4.5
anthropic/claude-haiku-4.5
TextImage
Text
200K—$1.00$0.9$5.00$4.50$0.1$0.09$1.25$1.1310% OFF
Claude Opus 5
anthropic/claude-opus-5
TextImage
Text
1M—$5.00$4.50$25.00$22.50$0.5$0.45$6.25$5.6310% OFF
Claude Sonnet 5
anthropic/claude-sonnet-5
TextImage
Text
1M—$2.00$1.80$10.00$9.00$0.2$0.18$2.50$2.2510% OFF
Deepseek V4 Flash
deepseek/deepseek-v4-flash
Text
Text
1.05M—$0.44$0.396$1.32$1.19$0.014$0.0126$010% OFF
Deepseek V4 Flash Vision
deepseek/deepseek-v4-flash-vision
TextImage
Text
1.05M—$0.44$1.32$0.014$0—

Context is the supported model window. Input and output modalities include text, image, audio, and video. Successful responses are billed from token usage.

Stable model aliases

Latest Model IDs that always follow the newest release

Call the newest LLM in a model family through one stable alias without updating your client for every version.

Start the ID with ~

Set request field model to ~provider/series-latest. The leading ~ is required.

Automatically follow the latest release

The gateway routes to the newest concrete model in the family. When a new version ships, the alias changes target and your client stays unchanged.

Billed at the current target rate

Pricing and capabilities follow the current target. For evaluations or regression tests, use a fixed model ID such as openai/gpt-5.6-sol.

Latest Model IDCurrently routes to
~openai/gpt-latestopenai/gpt-6-astra
~anthropic/claude-fable-latestanthropic/claude-fable-5.1
~anthropic/claude-haiku-latestanthropic/claude-haiku-4.5
~anthropic/claude-opus-latestanthropic/claude-opus-5
~anthropic/claude-sonnet-latestanthropic/claude-sonnet-5
~anthropic/sonnet-latestanthropic/claude-sonnet-5
~google/gemini-flash-latestgoogle/gemini-3.8-flash
~google/gemini-pro-latestgoogle/gemini-3.1-pro
~xai/grok-latestxai/grok-4.6
~moonshot/kimi-latestmoonshot/kimi-k3
~qwen/qwen-latestqwen/qwen3.8-max
~zai/glm-latestzai/glm-5.3

Targets change as each family evolves. The mappings on this page are current.

Client integrations

Supported clients

Point your existing OpenAI or Anthropic configuration at Modellix without changing the application protocol.

See the LLM API guide for the complete list.

Help

Frequently asked questions

Direct answers about developer integration and billing.

What is Modellix LLM?

Modellix LLM is a text-model gateway hosted at https://llm.modellix.ai. With one Modellix API key, you can call OpenAI-compatible Chat Completions and Responses or Anthropic-compatible Messages. Requests are synchronous with optional SSE streaming.

How is the LLM gateway different from the image and video API?

LLM text requests use https://llm.modellix.ai and return synchronously. Image, video, and speech generation use the asynchronous media API at https://api.modellix.ai. Do not mix the hosts or request fields.

How do I authenticate?

Create a key on the API Key page and send it as Authorization: Bearer or x-api-key. Do not use a model-provider platform key. The headers are equivalent; if both are present, they must contain the same key.

How should I format a model ID?

Pass provider/name in the request body, such as openai/gpt-5.6-sol, anthropic/claude-sonnet-5, or google/gemini-3.6-flash. Refer to product documentation for current models and prices.

Why do some models have pricing tiers?

Some models are priced by the request’s input-token count (Input Context). Tier count varies and is separate from the model context window. Context in the table is the supported window; Input Context is the pricing tier. Claude, most Gemini Flash, Qwen, and DeepSeek models are flat-rate. The calculator selects the tier automatically.

Which should I use: Chat Completions, Responses, or Messages?

OpenAI SDK, Codex, and Cursor use Chat Completions or Responses with https://llm.modellix.ai/v1. Anthropic SDK and Claude Code use Messages with https://llm.modellix.ai without /v1. The protocols have different URLs and request bodies; do not mix their fields.

How does billing work?

Successful responses are billed from token usage in USD per 1M tokens. Availability and prices may change; refer to product documentation. Insufficient balance returns 402.

Modellix LLM

Compatible with OpenAI and Anthropic protocols

Modellix does not require a separate SDK. Create an API key and point your existing client base URL at Modellix to call LLMs from Cursor, Claude Code, or your backend.

Get an API keyOpen ConsoleView the full LLM documentation