This platform is fully compatible with the OpenAI API format. You can use the official OpenAI SDKs (Python / Node.js / Java / Go, etc.) directly — just change the Base URL and API Key — or send requests with any HTTP library in any language.
Tip: All requests use Bearer Token authentication. Create your API Key on the "API Tokens" page in the console and keep it safe — never share it with anyone.
Authentication
Send your API Key in the Header of every request:
Authorization: Bearer sk-ah-你的密钥
Base URL
Set the following URL as the OpenAI SDK base_url:
https://www.ldyjs.cn/v1.php
Note: The Base URL must end with /v1.php — do not append /v1. The SDK automatically appends paths such as /chat/completions,/models and so on.
Available Models
Choose the endpoint by type: Text → POST /v1.php/chat/completions; Image → POST /v1.php/images/generations; Video → POST /video.php/videos(async task; then query the result with GET /video.php/videos/{id} ). Price units in the table below: text per 1M tokens, image per image, video per second (or per clip), as annotated in the table.
Product ID (model param)
Model
Brand
Type
Context Length
Input Price
Output Price
s9kxcc17lb1a
deepseek-v4-pro-0428
DeepSeek
Text
4,096
$0.7463/1M tokens
$3.7313/1M tokens
xl1srglahwx1
deepseek-v4-flash
DeepSeek
Text
4,096
$0.0245/1M tokens
$0.045/1M tokens
2p95hpgqniln
qwen3-30b-a3b
Qwen
Text
4,096
$0.0491/1M tokens
$0.184/1M tokens
eg3cmklgez2k
openpangu-2.0-pro
Huawei Pangu
Text
4,096
$0.1125/1M tokens
$0.3885/1M tokens
48wu00r8hpxs
qwen3-32b
Qwen
Text
4,096
$0.0491/1M tokens
$0.184/1M tokens
fz4uumurq0ls
openpangu-2.0-flash
Huawei Pangu
Text
4,096
$0.0204/1M tokens
$0.0409/1M tokens
0so8v1zcu2v0
deepseek-v4-pro
DeepSeek
Text
4,096
$0.2862/1M tokens
$0.5725/1M tokens
akkf2eay2ooi
glm-5.2
Zhipu GLM
Text
4,096
$0.2045/1M tokens
$0.6543/1M tokens
d2s9mea3ofvw
kimi-k2.6
Moonshot AI
Text
4,096
$0.1533/1M tokens
$0.6134/1M tokens
m50uyl25l7vz
glm-5.3
Zhipu GLM
Text
4,096
$0.8955/1M tokens
$3.7313/1M tokens
bificxygqykk
doubao-seedance-2.0-260128
Doubao
Video
-
On demand
On demand
jezqjbcp572b
doubao-seedance-2.0-fast-260128
Doubao
Video
-
On demand
On demand
l4fu7l8dy8ch
doubao-seedance-2.0-mini-260615
Doubao
Video
-
On demand
On demand
sm3t091meunu
doubao-seedance-2.5-260628
Doubao
Video
-
On demand
On demand
lsdfph68gehw
doubao-seedream-5.0-lite
Doubao
Image
-
$0.0224/image
$0.0224/image
0ndelflcyf86
doubao-seedream-5.0-pro-260628
Doubao
Image
-
$0.0021/image
$0.0313/image
zlxxdlzljqfy
doubao-seedream-5.0-flash-260915
Doubao
Image
-
On demand
$0.0125/image
5xr1fwnx8gno
seedance-2.5 5.8元 720P 30秒
Other
Video
-
On demand
$0.8657/item
Chat Completions
POST /chat/completions — Create a chat completion. The examples below cover six mainstream languages; click the tabs to switch:
GET /models — List thetextmodels accessible to the current token (the returned id is the Product ID and can be passed directly as the model parameter). Image/video products are not returned by this endpoint; refer to the "Available Models" table above.
from openai import OpenAI
client = OpenAI(
api_key="sk-ah-你的密钥",
base_url="https://www.ldyjs.cn/v1.php"
)
for m in client.models.list().data:
print(m.id)
import OpenAI from 'openai';
const client = new OpenAI({
apiKey: 'sk-ah-你的密钥',
baseURL: 'https://www.ldyjs.cn/v1.php'
});
const list = await client.models.list();
for (const m of list.data) console.log(m.id);
The token has no access to this model, or the IP is restricted
402
insufficient_quota
Insufficient quota — please top up or use another token
404
invalid_request_error
The model does not exist or has been delisted
502
upstream_error
The upstream service request failed
503
service_unavailable
No available channel for this model
SDKs & Dependencies
The platform is fully compatible with the OpenAI API — the official OpenAI SDKs are recommended. Java / Go / PHP can also use native HTTP libraries (all examples above are dependency-free). SDK installation per language:
Language
Installation
Description
Python
pip install openai
Official OpenAI SDK — just change base_url
Node.js
npm install openai
Official OpenAI SDK — just change baseURL
Java
com.squareup.okhttp3:okhttp
Native OkHttp calls; the official OpenAI Java SDK also works
Golang
Nothing to install
Standard library net/http, no extra dependencies
PHP
composer require guzzlehttp/guzzle
Guzzle HTTP client; openai-php/client also works
cURL
Built into the OS
Call directly from the command line — great for quick testing
Tip: Whatever language you use, only two substitutions are needed: base_url to https://www.ldyjs.cn/v1.php, and your API Key to the token created in the console.
Billing
Fees are calculated based on actual token usage. Formula:
Image models: billed per image (tokens in the formula above translate to image count);Video models: billed per second or per clip; see the "Video Generation" page in the console for cost estimates
The usage field in every response returns the exact token usage for cost calculation. Non-streaming responses always include usage; for streaming responses it depends on the upstream — some providers include it in the last chunk.
Token quota: Each token can have its own quota. Requests check both the account balance and the token balance; if either is zero, a 402 error is returned.