GPT-6 Luna

GPT-6 Luna is OpenAI’s most efficient GPT-6 model for focused, high-volume tasks. Use the model ID gpt-6-luna with the OpenAI-compatible Responses API at POST /v1/responses or Chat Completions at POST /v1/chat/completions.

Prefer the Responses API for built-in tools, function calling with reasoning, image input, Structured Outputs, and prompt caching. Chat Completions function calling is supported only when reasoning_effort is none.

Model specifications

Property	Value
Model ID	`gpt-6-luna`
Recommended API	Responses API
Input	Text and image
Output	Text
Context window (official specification)	1,050,000 tokens
Maximum output	128,000 tokens
Knowledge cutoff	May 18, 2026
Reasoning effort	`none`, `low`, `medium`, `high`, `xhigh`, `max`
Default reasoning effort	`medium`
Fine-tuning	Not supported

When to use GPT-6 Luna

Focused, high-volume text processing
Classification, extraction, and structured output
Cost- and throughput-sensitive agent workflows
Large-context summarization and analysis
Vision tasks that combine text and image input
Routine coding and automation tasks

Quick start

Responses API
Chat Completions

cURL

curl https://www.anyfast.ai/v1/responses \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-6-luna",
    "input": "Classify this support ticket and return a short routing reason.",
    "reasoning": {
      "effort": "medium",
      "summary": "auto"
    },
    "max_output_tokens": 1024
  }'

cURL

curl https://www.anyfast.ai/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-6-luna",
    "messages": [
      { "role": "user", "content": "Summarize this incident report in three bullets." }
    ],
    "reasoning_effort": "medium",
    "max_completion_tokens": 1024
  }'

Tool calling

Use the Responses API when the model needs to reason and call tools:

cURL

curl https://www.anyfast.ai/v1/responses \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-6-luna",
    "input": "Look up the order status for order AF-1024.",
    "reasoning": { "effort": "medium" },
    "tools": [{
      "type": "function",
      "name": "get_order_status",
      "description": "Return the current status of an order.",
      "parameters": {
        "type": "object",
        "properties": {
          "order_id": { "type": "string" }
        },
        "required": ["order_id"],
        "additionalProperties": false
      },
      "strict": true
    }],
    "tool_choice": "auto"
  }'

In Chat Completions, function calling requires reasoning_effort: "none". Use the Responses API if you need tool calling with low, medium, high, xhigh, or max reasoning.

Chat Completions wraps each function definition inside a function object:

{
  "type": "function",
  "function": {
    "name": "get_order_status",
    "description": "Return the current status of an order.",
    "parameters": {
      "type": "object",
      "properties": {
        "order_id": { "type": "string" }
      },
      "required": ["order_id"],
      "additionalProperties": false
    },
    "strict": true
  }
}

Built-in tools

OpenAI lists the following Responses API tools for GPT-6 Luna: Web search, File search, Image generation, Code Interpreter, Hosted shell, Apply patch, Skills, Computer use, MCP, and Tool search. Each tool has its own request fields and may require supporting resources such as files, vector stores, remote servers, or a computer-use loop.

Reasoning and sampling rules

Effort	Typical use
`none`	Lowest latency, sampling controls, or Chat Completions function calling
`low`	Routine tasks with light reasoning
`medium`	General analysis and automation; the default
`high`	Complex multi-step work
`xhigh`	Difficult professional tasks
`max`	Maximum reasoning depth

When reasoning effort is not none, omit temperature, top_p, and top_logprobs. For Chat Completions, also omit logprobs. For Responses, do not request message.output_text.logprobs in include.

Core parameters

Parameter	API	Description
`model`	Both	Must be `gpt-6-luna`
`input`	Responses	Text or ordered input items
`messages`	Chat Completions	Ordered conversation messages
`reasoning.effort`	Responses	`none`, `low`, `medium`, `high`, `xhigh`, or `max`
`reasoning_effort`	Chat Completions	Same effort values as Responses
`max_output_tokens`	Responses	Maximum visible output and reasoning tokens, up to 128,000
`max_completion_tokens`	Chat Completions	Maximum completion and reasoning tokens
`tools`	Both	Tool definitions; Chat Completions requires `reasoning_effort: "none"`
`text.format`	Responses	Text or strict JSON Schema output configuration
`stream`	Both	Stream response events

In raw Responses API JSON, read generated text from message items in output[].content[].text. Some official SDKs expose an aggregated output_text convenience property, but clients should not require a top-level output_text field in the raw HTTP response.

API Reference

View the GPT-6 Luna API reference.

Getting started

Model Introduction

Resources

Model specifications

When to use GPT-6 Luna

Quick start

Tool calling

Built-in tools

Reasoning and sampling rules

Core parameters

API Reference

Official references

​Model specifications

​When to use GPT-6 Luna

​Quick start

​Tool calling

​Built-in tools

​Reasoning and sampling rules

​Core parameters

API Reference

​Official references

Model specifications

When to use GPT-6 Luna

Quick start

Tool calling

Built-in tools

Reasoning and sampling rules

Core parameters

Official references