DINGJIAI
DINGJIAI
Quick start in 5 minutesWhat is DINGJIAI?

Online use

Online chatOpens in a new windowImage generationOpens in a new window

Popular Agents

Codex SetupClaude Code SetupOne-Click Setup

API Reference

API overview
Base URL and authenticationHeadersModel lists and model selection
Error codes

Billing

Model pricingTroubleshooting calls
Request basics

Model lists and model selection

Find models published for the current API entry and choose the correct public model id.

Model products are isolated by API entry

An administrator may publish the same public model id to multiple internal entries as separate products. The model value or path model name must be available for the current entry; one entry's model list cannot prove another entry is available.

Query model lists

OpenAI Chat Completions

GET /v1/models returns products published for the Chat Completions entry.

Anthropic

GET /anthropic/v1/models returns the Anthropic-style model list.

Gemini

GET /v1beta/models returns the Gemini-style model list.

GET /v1/models
curl https://dingjiai.com/v1/models \
  -H "Authorization: Bearer YOUR_USER_API_KEY"
GET /anthropic/v1/models
curl https://dingjiai.com/anthropic/v1/models \
  -H "x-api-key: YOUR_USER_API_KEY"
GET /v1beta/models
curl https://dingjiai.com/v1beta/models \
  -H "x-goog-api-key: YOUR_USER_API_KEY"

What the lists contain

Model lists return products that are publicly visible, enabled, and have an available upstream candidate. They are not complete native capability matrices; a real call checks the current entry, active product, balance, concurrency, and upstream candidates again.

Responses, Codex Search, and Images Generations use public model ids explicitly enabled for their corresponding entries and currently have no dedicated public model list. The web-only openai_images_edits operation is not a public API option.

Selection order

  1. Choose the API family and entry first. Chat Completions uses the OpenAI list, Anthropic uses the Anthropic list, and Gemini uses the Gemini list.
  2. Copy the product's public model id. Do not guess names or use an upstream model name unless DINGJIAI publishes that exact id.
  3. Send a minimal real request. The gateway checks balance, concurrency, and upstream availability again.

The console's /api/models endpoint is for the signed-in console UI, not a client integration model-list API. GET /v1/models only proves OpenAI Chat Completions products are visible; it does not prove Responses or Images availability.

Last updated on

HeadersResponses

On this page

Query model listsWhat the lists containSelection order