gpt-6-astra API
GPT-6 Astra is an OpenAI model for complex reasoning, coding and research, with text and image input and text output. Use XiuAI to check its exact model ID, current prices and API protocols.[1][2]
gpt-6-astraModel overview
Published provider specifications. Available features depend on the current API route.
- API routes
- OpenAI-compatible API / openai-response / openai-response-compact / Anthropic Messages / Gemini GenerateContent / openai-alpha-search
Where GPT-6 Astra fits
Consider Astra when a task involves several constraints, code and source material, or research that must become a reviewable document. State the goal, available evidence and completion criteria. Ask it to preserve supporting evidence and unresolved questions, then verify code, facts and calculations in your application.[1]
Astra accepts images for analysis; this does not make image generation its direct output modality. For fixed-format conversion or simple classification, compare task samples, current prices and required capabilities across models instead of choosing solely by flagship positioning.[1][2]
Reasoning effort and tool boundaries
When setting reasoning.effort, check the selected protocol and its allowed values. A higher setting can change latency and usage without guaranteeing a correct answer. Choose the setting that meets your task’s acceptance criteria.[1]
Computer use, search, file retrieval and hosted execution in OpenAI’s documentation are specific tools or platform features. Check the protocol and tool fields used through XiuAI separately. Suitability for agent tasks does not mean XiuAI hosts those tools or your business workflow.[1][2]
Comparing request costs
Compare input, output, cache reads and cache writes using the current rates and conditions on this page. Long-context requests can fall into a different pricing tier; the context limit and the pricing threshold are separate. The same question can produce different output lengths, so reconcile charges in request records.[1][2]
API access
Using Astra through XiuAI
Use gpt-6-astra with your existing XiuAI account, balance and an authorized API key. Select a protocol listed in the current catalog and use its request structure. Responses API input and Chat Completions messages are not interchangeable.[1][2]
Start with a short request that represents the real task, then add source material or tools after checking the output and parameters. Use the catalog for current groups, protocols and billing conditions, and the vendor page for model information. Vendor quotas, tools and prices do not establish your XiuAI account configuration.[1][2]
Send a request using your existing API key.
POST /v1/chat/completionscurl 'https://router-api.xiu.ai/v1/chat/completions' \
-H "Authorization: Bearer $XIUAI_API_KEY" \
-H 'Content-Type: application/json' \
--data-raw '{
"model": "gpt-6-astra",
"messages": [
{
"role": "user",
"content": "Hello"
}
]
}'Read the API documentationFrequently asked questions
Does Astra directly generate images or video?
The model takes text and images and produces text. Image-generation tools in vendor documentation are a separate capability. Image input does not establish image or video generation through XiuAI; use an available media model for that task.[1][2]
Is higher reasoning effort better for every task?
No. Define the constraints, accuracy requirements and time budget, then compare representative requests. The availability of a higher setting is not a reason to use it for a simple task, and complex results still need review.[1]
Why can long inputs use different rates from short requests?
Astra pricing includes context-tier conditions, and cache type and output usage also affect the total. Read the current pricing tiers and units, then reconcile the request record. Supported input length does not imply one rate for every request.[1][2]
Can I copy every OpenAI tool example unchanged?
First check whether the example’s protocol and hosted tools apply to the current integration. A model request, functions implemented by your application and vendor-hosted search or execution are different capabilities. Start with the current XiuAI example and add only the fields you need.[1][2]
Related models
gpt-6.1-solgpt-6-lunaSources and updates
Specifications reference provider documentation; prices and availability come from the current XiuAI catalog.
Content reviewed: 2026-11-16
[1] OpenAI GPT-6 Astra documentation[2] Current XiuAI model and pricing dataModel documentationTool callingStructured outputContext documentation