Supported Models

We2AI exposes the following LLMs through an OpenAI-compatible API. Change base_url to https://we2ai.ai/v1, use your We2AI token, and choose the model name you need.

DeepSeek · DeepSeek

DeepSeek is strong in reasoning, coding and Chinese-language work. DeepSeek V4 is suitable for general chat and coding tasks, while DeepSeek-R1 is designed for heavier reasoning and debugging workloads.

• Use cases: chat, code generation, document analysis, mathematical reasoning
• Context: up to 128K tokens depending on model
• Model names: deepseek-chat, deepseek-reasoner
• Clients: Cursor, Continue, Cline, LobeChat, NextChat, ChatBox

Qwen · Alibaba Cloud

Qwen covers general chat, coding, multimodal and long-context workloads. It is a practical default for Chinese instruction following, tool calling, agents and RAG systems.

• Use cases: Chinese chat, agents, function calling, code generation, enterprise RAG
• Context: 128K to 1M tokens depending on model
• Model names: qwen-plus, qwen-max, qwen-coder-plus
• Clients: OpenAI-compatible clients

GLM · Zhipu AI

GLM models are useful for bilingual chat, structured JSON output and lower-cost batch classification or extraction tasks. GLM-4.5 is the flagship line; GLM-4-flash is suitable for high-volume low-cost jobs.

• Use cases: bilingual chat, structured output, classification, extraction
• Context: up to 128K tokens
• Model names: glm-4.5, glm-4.5-air, glm-4-flash
• Clients: OpenAI-compatible clients

Kimi · Moonshot AI

Kimi is known for long-context document processing. It is a good choice when you need to analyze books, reports, large files or repository-scale context.

• Use cases: long document analysis, research summaries, repository QA
• Context: 128K to 1M tokens depending on model
• Model names: moonshot-v1-128k, moonshot-v1-1m, kimi-k1
• Clients: OpenAI-compatible clients

How to Call

Python

from openai import OpenAI

client = OpenAI(
    base_url="https://we2ai.ai/v1",
    api_key="sk-YOUR_WE2AI_TOKEN",
)

resp = client.chat.completions.create(
    model="deepseek-chat",
    messages=[{"role": "user", "content": "Introduce yourself briefly"}],
)
print(resp.choices[0].message.content)

cURL

curl https://we2ai.ai/v1/chat/completions \
  -H "Authorization: Bearer sk-YOUR_WE2AI_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen-plus",
    "messages": [{"role":"user","content":"hi"}]
  }'

Node.js

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://we2ai.ai/v1",
  apiKey: process.env.WE2AI_TOKEN,
});

const r = await client.chat.completions.create({
  model: "glm-4.5",
  messages: [{ role: "user", content: "hi" }],
});
console.log(r.choices[0].message.content);