deepseek-v4-flashV4 default — 1M context, 384k max output.
Multi-turn dialogue, system prompts.
SSE chunks for incremental output.
Structured arguments via JSON schemas.
Pass images alongside the prompt.
Text and image to short video, async.
Coming soon.
This example calls the selected DeepSeek route with deepseek-v4-flash.
import OpenAI from 'openai';
const client = new OpenAI({
baseURL: 'https://api.miavo.xyz/v1',
apiKey: process.env.MACAW_API_KEY!,
});
const res = await client.chat.completions.create({
model: 'deepseek-v4-flash',
messages: [
{ role: 'user', content: 'Write me a haiku about gateways.' },
],
});
console.log(res.choices[0].message.content);