Build with
signal speed.

import { Axon } from '@axon/sdk';

const client = new Axon({
  apiKey: process.env.AXON_API_KEY,
  region: 'us-west-2',
});

// Generate a completion — P95 <20ms guaranteed
const completion = await client.inference.complete({
  model: 'llama-3-70b',
  messages: [{ role: 'user', content: 'Explain neural networks' }],
  max_tokens: 512,
});

console.log(completion.choices[0].message.content);