OpenAI
GPT-6 Luna is an efficient OpenAI model for focused, high-volume tasks, with image input, structured outputs, and configurable reasoning.
At a glance
API model ID: gpt-6-luna
Capabilities & boundaries
Tool support requires the appropriate API integration; a supported tool is not automatically active in every chat.
Medium is the documented API default. For a tightly scoped task, compare none or low against medium on examples with known answers. Increase effort when the task needs more reasoning, and measure correctness, response time, and token use. None disables reasoning; minimal is not a supported setting. These API controls may not all be exposed in the app.
Unsupported settings: minimal.
Put it to work
Original examples from EZ Ai Assist. Adapt these to your task and the features available in your workspace.
Workflow 01
Supply the allowed labels, routing rules, and examples so classification stays consistent and reviewable.
Classify each request below using only the labels and routing rules I provide. Return its request ID, chosen label, a short supporting quote, and a brief explanation. Use needs review if the evidence is ambiguous or no label fits. Do not invent facts, add labels, or route requests automatically. Flag cases where the rules conflict.
Workflow 02
Give the model the source material and an explicit schema, then validate the result before using it.
Extract data from the supplied records into the JSON schema below. Preserve record IDs and copy values faithfully. Use null for missing or uncertain fields where the schema permits it; never guess. Include source evidence only in fields defined by the schema. Flag any conflict between the source and the schema instead of silently changing the data. Return no extra fields.
Workflow 03
Use approved policies and account facts to prepare a concise reply for a person to review.
Draft a concise, helpful reply to the customer message below using only the approved policy and account facts I provide. Acknowledge the issue, explain the supported next step, and ask one clarifying question if needed. Do not invent refunds, timelines, account actions, or promises. Put missing facts and policy conflicts in a separate review note. Do not send the reply or take account actions.
Developer reference
These are OpenAI API reference prices, not EZ Ai Assist subscription prices.
View EZ Ai Assist plans| Token type | ≤ 272,000input tokens | > 272,000input tokens |
|---|---|---|
| Input | $0.10 | $0.20 |
| Cached input | $0.01 | $0.02 |
| Cache writes | $0.125 | $0.25 |
| Output | $0.50 | $0.75 |
Common questions
OpenAI positions Luna for focused, high-volume work. Try it on a representative set of requests with a clear rubric and include the cost of human review. Compare Sol or Astra when tasks have difficult dependencies, conflicting evidence, or a high cost of error. Lower token prices alone do not establish the best total cost or result.
Medium is Luna’s documented API default. None disables reasoning and is worth testing for narrowly defined tasks; low through max provide reasoning options. Compare results with known answers instead of assuming one setting is always best. The app may expose a different set of controls from the direct API.
Yes, but only with reasoning_effort set to none. Use the Responses API for function calling with reasoning and for built-in tools. Tool support also requires an integration to supply the tools and handle permissions; it does not mean those tools are enabled in every EZ Ai Assist conversation.
No. A prompt alone is not a guarantee of schema-compliant or factually correct data. In a direct API integration, configure structured outputs with a supported schema, validate the response, and handle refusals, incomplete responses, and missing evidence. Check which output controls the app supports before building a workflow around them.
Include relevant material with clear source names and ask for evidence you can check. A larger window allows more input; it does not guarantee that every detail will be used correctly. The maximum output limit is separate, and input above the pricing threshold changes the rates for the entire request.
Yes, as editorial starting points. Supply your labels, schema, policies, or records and include only information you are authorized to share. Review the output before routing requests, updating records, or sending replies. The examples are not benchmark results or guarantees of a particular outcome.
No. The table describes direct OpenAI API usage, including separate long-context rates. EZ Ai Assist subscriptions and available model controls are separate. Visit our pricing page for current plans and check the app for supported options.
Check the source
Specifications and API prices checked on . Example prompts and workflow advice are editorial guidance from EZ Ai Assist.
Same provider
OpenAI
GPT-6.1 Sol is OpenAI’s reasoning model for complex coding and professional work, with image input and a 1.05M-token context window.
OpenAI
GPT-6 Sol is an OpenAI model for complex coding and tool-assisted workflows, with image input and configurable reasoning.
OpenAI
GPT-6 Astra is an OpenAI model for complex reasoning, coding, and research, with image input and a 1.05M-token context window.
OpenAI
GPT-5.6 Sol is an OpenAI model for professional reasoning and coding, with image input, configurable reasoning effort, and a 1.05M-token context window.