List available models

GEThttps://api.poe.com/v1/models

Overview

Returns a list of all models available via the Poe API.

Popular models include:

  • GPT-5-Pro - OpenAI's latest flagship model with extended reasoning
  • GPT-5-Codex - Specialized for software engineering tasks
  • Claude-Sonnet-4.5 - Anthropic's most advanced model
  • Claude-Haiku-4.5 - Fast and efficient with frontier intelligence
  • Gemini-2.5-Pro - Google's advanced model with web search
  • Grok-4 - xAI's most intelligent language model
  • DeepSeek-R1 - Top open-source reasoning model

You can also access specialized bots for image generation (Nano-Banana, Amazon-Nova-Canvas), video generation (Amazon-Nova-Reel-1.1), and many other community-created models.

Note: This endpoint does not require authentication and returns all publicly available models.

Authentication

Send your Poe API key in the Authorization header:

Authorization: Bearer sk_test_51SAMPLEKEY

All requests must be made over HTTPS.

Parameters

This endpoint does not accept query or path parameters.

Responses

FieldTypeRequiredDescription
object"list"OptionalObject type identifier Allowed values: list
dataobject[]Optional
data[].idstringRequiredModel identifier
data[].object"model"RequiredObject type, always "model" Allowed values: model
data[].createdint64RequiredUnix timestamp in milliseconds of when the model was created
data[].descriptionstringOptionalDetailed description of the model's capabilities, features, and usage instructions
data[].owned_bystringRequiredOrganization that owns the model
data[].permissionobject[]OptionalPermission information for the model (typically empty)
data[].rootstringOptionalRoot model identifier
data[].parentstring | nullOptionalParent model identifier if this is a derivative model
data[].architectureobjectOptionalModel architecture and modality information
data[].architecture.input_modalities"text" | "image" | "video" | "audio"[]OptionalSupported input modalities
data[].architecture.output_modalities"text" | "image" | "video"[]OptionalSupported output modalities
data[].architecture.modalitystringOptionalSimplified modality string representation
data[].pricingobject | nullOptionalPricing information for the model. Null for free models or models with request-based pricing

❌ Error codes

HttpTypeDescription
429rate_limit_errorRate limit exceeded Rate limit exceeded (500 requests per minute)

Best Practices

Cache the models list in your application. Model availability doesn't change frequently, so you can safely cache this response for 1-24 hours.

This endpoint returns all publicly available models. Consider:

  • Caching: Models list changes infrequently, cache for 1-24 hours
  • Filtering: Filter models by architecture.input_modalities for your use case
  • Pricing: Use the pricing field to estimate costs before making requests

🔁 Callbacks & webhooks

No callbacks or webhooks are associated with this endpoint.