Skip to content
Docs menu

LLM models

GPT-5.6 Terra

On this page

GPT-5.6 — frontier reasoning and generation through one API.

modelIdgpt-5-6-terra
Modalitytext
PricingSee this model on the Pricing page for the current per-call price (with your markup).

Operations

OperationmodelIdEndpointRequired input
Defaultgpt-5-6-terraPOST /api/v1/generateprompt
Poll taskGET /api/v1/task/{id}?model=gpt-5-6-terra

Input parameters

FieldTypeRequiredValues / example
promptstringYesText value (example: Compare two ways to store user files for a small API: object storage with signed URLs, versus streaming through our own server. For each, give the failure mode… — full value in the request example)
memorybooleanNoAutomatically append previous messages to maintain multi-turn context. May increase token usage. (true/false) (default: false)
web_searchbooleanNoWeb search feeds the results back as input tokens, so a searched turn typically costs many times a normal turn (measured: 18–87×). Billed per token as usual — nothing extra per search. (true/false) (default: true)
reasoning_effortstringNoControl reasoning depth: Low for faster responses, High for deeper analysis (options: Low | Medium | High | XHigh) (default: Medium)
streambooleanNoSend true to receive the reply as an event stream (Content-Type: text/event-stream). Omit it and you get a single JSON response — streaming is opt-in on this endpoint. (true/false) (example: true)
imagesstring[]NoUp to 4 images per message — JPEG, PNG, GIF or WebP, 5 MB each. Send each one as a data: URL or an https URL. Images are billed as input tokens. (image URL)
systemstringNoSystem instruction for the model. Defaults to "You are GPT-5.6 Terra, developed by OpenAI." when omitted.
messagesarrayNoA conversation instead of a single `prompt`: up to 20 objects of { "role", "content" }, where role is user, assistant or system. Send `prompt` or `messages`, not both. Do not flatten a conversation into one prompt with "User:" / "Assistant:" labels — the model reads that as a single message, and a model using web search then searches for the whole block instead of your question.

Example request

bash
curl -X POST https://you.bot/api/v1/generate \
  -H "Authorization: Bearer $YOUBOT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"modelId":"gpt-5-6-terra","input":{"prompt":"Compare two ways to store user files for a small API: object storage with signed URLs, versus streaming through our own server. For each, give the failure mode that bites first at scale, the cost that grows fastest, and one thing that is hard to change later. Then pick one for a service serving 10,000 files a day and say what would change your mind. Under 250 words.","memory":false,"web_search":true,"reasoning_effort":"Medium","stream":true,"images":["data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAIAAAACCAIAAAD91JpzAAAAFklEQVR4nGP8z4AATAxIHFQeMg8OAgB1yQIDpk8YwwAAAABJRU5ErkJggg=="]}}'