Streaming - OpenModex

Vue d’ensemble

Définissez stream: true pour recevoir les réponses sous forme de Server-Sent Events (SSE). Chaque événement contient un fragment de réponse partiel, vous permettant d’afficher le contenu en temps réel au fur et à mesure de sa génération.

Format du flux

Le flux envoie des lignes data: contenant des objets JSON :

data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"role":"assistant"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"content":"Hello"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"content":"!"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}

data: [DONE]

Objet chunk

Chaque chunk contient :

Champ	Type	Description
`id`	string	Partagé entre tous les chunks d’une même réponse
`object`	string	Toujours `chat.completion.chunk`
`created`	integer	Horodatage Unix
`model`	string	Modèle ayant généré la réponse
`choices[].delta.role`	string	Présent uniquement dans le premier chunk
`choices[].delta.content`	string	Contenu du token (peut être vide)
`choices[].finish_reason`	string \| null	`stop`, `length` ou `null` pendant le streaming

Exemples

curl https://api.openmodex.com/v1/chat/completions \
  -H "Authorization: Bearer omx_sk_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o",
    "messages": [{"role": "user", "content": "Write a poem about the sea."}],
    "stream": true
  }'

import OpenModex from '@openmodex/sdk';

const client = new OpenModex({ apiKey: 'omx_sk_YOUR_KEY' });

const stream = await client.chat.completions.create({
  model: 'gpt-4o',
  messages: [{ role: 'user', content: 'Write a poem about the sea.' }],
  stream: true,
});

for await (const chunk of stream) {
  const content = chunk.choices[0]?.delta?.content;
  if (content) process.stdout.write(content);
}

from openmodex import OpenModex

client = OpenModex(api_key="omx_sk_YOUR_KEY")

stream = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Write a poem about the sea."}],
    stream=True,
)

for chunk in stream:
    content = chunk.choices[0].delta.content
    if content:
        print(content, end="", flush=True)

import { useChat } from '@openmodex/react';

function Chat() {
  const { messages, input, setInput, sendMessage, isLoading, stop } = useChat({
    apiKey: 'omx_sk_YOUR_KEY',
    model: 'gpt-4o',
  });

  return (
    <div>
      {messages.map((m) => (
        <p key={m.id}><b>{m.role}:</b> {m.content}</p>
      ))}
      <input value={input} onChange={(e) => setInput(e.target.value)} />
      <button onClick={() => sendMessage()} disabled={isLoading}>Send</button>
      {isLoading && <button onClick={stop}>Stop</button>}
    </div>
  );
}

Annuler un flux

Vous pouvez annuler un flux en cours :

const stream = await client.chat.completions.create({
  model: 'gpt-4o',
  messages: [{ role: 'user', content: 'Write a long story...' }],
  stream: true,
});

// Abort after 2 seconds
setTimeout(() => stream.abort(), 2000);

for await (const chunk of stream) {
  process.stdout.write(chunk.choices[0]?.delta?.content ?? '');
}

stream = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Write a long story..."}],
    stream=True,
)

for chunk in stream:
    content = chunk.choices[0].delta.content
    if content:
        print(content, end="", flush=True)
    # Call stream.close() to abort