Get the Model List of AceDataCloud Platform

Returns all available LLM / image / video / audio / embedding / reordering models on the platform — returns in a flat view compatible with OpenAI /v1/models style, compatible with OpenAI SDK. A Proxy service typically corresponds to dozens or even hundreds of models.

Applicable scenarios:

Render "model selection dropdown" in your product.
Perform model availability checks at the routing layer.
Sync to LiteLLM / OpenRouter style middleware.

ℹ️ This interface belongs to the AceDataCloud Platform Management API, with a unified prefix of https://platform.acedata.cloud/api/v1/.

¶ Interface Overview

Item	Content
Method	`GET`
URL	`https://platform.acedata.cloud/api/v1/models/`
Authentication	❌ Public

¶ Authentication Description

This interface is completely public, no account token or login state required.

💡 However: if an account token is included, the response will be filtered based on the services you have applied for, returning only the models you can call.

¶ Query Parameters

Parameter	Type	Required	Default	Description
`service_id`	UUID	No	—	Filter by service
`tag`	string	No	—	Filter by tag (e.g., `chat`, `embedding`, `image`, `video`, `tts`)
`limit`	integer	No	100	Number of items per page, maximum 200
`offset`	integer	No	0	Offset
`ordering`	string	No	—	Sorting field

¶ Request Examples

¶ cURL

# All models
curl 'https://platform.acedata.cloud/api/v1/models/?limit=200' \
  -H 'accept: application/json'

# Only view chat types
curl 'https://platform.acedata.cloud/api/v1/models/?tag=chat&limit=200' \
  -H 'accept: application/json'

¶ Python

import requests

resp = requests.get(
    "https://platform.acedata.cloud/api/v1/models/",
    headers={"accept": "application/json"},
    params={"tag": "chat", "limit": 200},
    timeout=10,
)
data = resp.json()
print(f"Total {data['count']} chat models")
for m in data["items"][:20]:
    print(f"  {m['id']:50s}  ${m.get('prompt_price') or 0:.6f}/1K in, ${m.get('completion_price') or 0:.6f}/1K out")

¶ Node.js

const r = await fetch('https://platform.acedata.cloud/api/v1/models/?limit=200')
const { count, items } = await r.json()
console.log(`Total ${count} models`)

¶ Response Example (HTTP 200)

{
  "count": 187,
  "items": [
    {
      "id": "gpt-4.1",
      "object": "model",
      "owned_by": "openai",
      "service_id": "38ecf158-36f2-42f2-8e7f-6786cdfc2452",
      "service_alias": "openai",
      "service_title": "OpenAI",
      "context_window": 1048576,
      "max_output_tokens": 32768,
      "prompt_price": 0.000002,
      "completion_price": 0.000008,
      "cached_prompt_price": 0.0000005,
      "tags": ["chat", "vision", "tool_call"],
      "created_at": 1733616000
    }
  ]
}

ℹ️ Field names are also compatible with OpenAI's official /v1/models response (id, object, owned_by), and can be directly fed to the OpenAI SDK.

¶ Response Field Description

Field	Type	Description
`id`	string	Model ID (e.g., `gpt-4.1`, `claude-sonnet-4`, `midjourney`, `veo-3`)
`object`	string	Always `"model"` (OpenAI compatible field)
`owned_by`	string	Owner of the model (`openai` / `anthropic` / `google` / `meta` / `xai`...)
`service_id`	UUID	The platform service ID to which this model is attached
`service_alias`	string	Service alias
`service_title`	string	Service title
`context_window`	integer	Context window length (Token count, specific to LLM)
`max_output_tokens`	integer	Maximum output Token count per request
`prompt_price`	number	null
`completion_price`	number	null
`cached_prompt_price`	number	null
`tags`	array	Model capability tags: `chat` / `vision` / `tool_call` / `audio` / `embedding` / `rerank` / `image` / `video` / `tts` / `stt` etc.
`created_at`	integer	Model launch time (Unix timestamp in seconds, OpenAI compatible field)

¶ Error Handling

HTTP	code	Meaning
400	`invalid`	Invalid parameter

¶ Practical Tips

OpenAI SDK Compatible: The response from this interface can be directly used as a replacement source for client.models.list() in the OpenAI Python SDK.
Prices are per Token: Not per thousand Tokens — this differs from the "per 1M tokens" shown in OpenAI's official pricing table. Remember to multiply by 1,000,000 when converting.
null price indicates billing by call count: Typically for image / video / audio models, each call is billed by a fixed number of Credits. See the package table in Creating AceDataCloud Platform Recharge Orders.
Filter actual callable models: Including an account token will return only the models under the services you have applied for.