Qwen OpenAI-compatible API

Test Qwen-style Chinese models through an OpenAI-compatible request shape.

Botat helps overseas developers evaluate Qwen, DeepSeek, Kimi, and other Chinese-strong models without rewriting apps that already speak OpenAI-style chat completions.

Create account and run a smoke test

Compatibility does not mean identical behavior

Request shape

Start with /v1/chat/completions, a bearer key, messages, model, and stream. Keep wrappers out of the first test.

Capability flags

Check response_format, tools, tool_choice, streaming deltas, stop, seed, penalties, and provider-specific fields separately.

Model mapping

Use explicit model IDs from your account and keep a fallback map in config instead of hard-coding one provider path.

Retry behavior

Measure p95 latency, 429/5xx rates, and retry safety with the same prompt set you expect to run in production.

Minimal app configuration

const client = new OpenAI({
  apiKey: process.env.BOTAT_API_KEY,
  baseURL: "https://botat.com/v1"
});

const result = await client.chat.completions.create({
  model: "YOUR_ACCOUNT_QWEN_MODEL_ID",
  messages: [{ role: "user", content: "Summarize this Chinese paragraph in English." }],
  stream: false
});

Keep the model ID as configuration so you can test Qwen-style, DeepSeek-style, and Kimi-style routes without changing application code.

Good-fit workloads

Non-sensitive Chinese-language summarization, localization QA, batch rewriting, coding assistance, internal text cleanup, and retryable async jobs are the cleanest fit.

Data boundary

Do not send sensitive personal data, confidential customer data, regulated data, or workloads with strict residency requirements through China-model routes. Keep this path for non-sensitive, retryable work.