API-Dokumentation

So funktioniert es

Die Seven Assistant API ist vollständig OpenAI-kompatibel. Das bedeutet: Jedes Tool, jede Bibliothek und jedes SDK, das mit der OpenAI API funktioniert, funktioniert auch mit Seven Assistant. Sie müssen nur die Base URL ändern.

Erste Anfrage mit curl

curl YOUR_DOMAIN/v1/chat/completions \
  -H "Authorization: Bearer sk-educa-YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "educa-ai",
    "messages": [
      {"role": "system", "content": "Du bist ein hilfreicher Assistent."},
      {"role": "user", "content": "Was ist Photosynthese?"}
    ]
  }'

Modelle per API abrufen

curl YOUR_DOMAIN/v1/models \
  -H "Authorization: Bearer sk-educa-YOUR_API_KEY"

So funktioniert Streaming

Setzen Sie "stream": true im Request. Die Antwort wird als Server-Sent Events (SSE) gestreamt, jeder Chunk enthält ein Textfragment. Der Stream endet mit data: [DONE].

curl YOUR_DOMAIN/v1/chat/completions \
  -H "Authorization: Bearer sk-educa-YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -N \
  -d '{
    "model": "educa-ai",
    "messages": [{"role": "user", "content": "Zähle bis 5"}],
    "stream": true
  }'

Stream-Chunks

data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","created":1711900000,"model":"educa-ai","choices":[{"index":0,"delta":{"role":"assistant"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","created":1711900000,"model":"educa-ai","choices":[{"index":0,"delta":{"content":"1, "},"finish_reason":null}]}

data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","created":1711900000,"model":"educa-ai","choices":[{"index":0,"delta":{"content":"2, "},"finish_reason":null}]}

data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","created":1711900000,"model":"educa-ai","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}

data: [DONE]

from openai import OpenAI

client = OpenAI(
    api_key="sk-educa-YOUR_API_KEY",
    base_url="YOUR_DOMAIN/v1/"
)

# Chat Completion
response = client.chat.completions.create(
    model="educa-ai",
    messages=[
        {"role": "system", "content": "Du bist ein hilfreicher Lehrer."},
        {"role": "user", "content": "Erstelle 5 Quizfragen zum Thema Mittelalter."}
    ],
    temperature=0.7,
    max_tokens=1000
)

print(response.choices[0].message.content)

# Streaming
stream = client.chat.completions.create(
    model="educa-ai",
    messages=[{"role": "user", "content": "Erkläre Demokratie."}],
    stream=True
)

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

# Embeddings
embedding = client.embeddings.create(
    model="educa-ai",
    input="Photosynthese erklärt für Kinder"
)

print(embedding.data[0].embedding[:5])

Installation: pip install openai

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: "sk-educa-YOUR_API_KEY",
  baseURL: "YOUR_DOMAIN/v1/"
});

// Chat Completion
const response = await client.chat.completions.create({
  model: "educa-ai",
  messages: [
    { role: "system", content: "Du bist ein hilfreicher Lehrer." },
    { role: "user", content: "Was ist der Dreisatz?" }
  ]
});

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

// Streaming
const stream = await client.chat.completions.create({
  model: "educa-ai",
  messages: [{ role: "user", content: "Erkläre den Wasserkreislauf." }],
  stream: true
});

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

Installation: npm install openai

const response = await fetch("YOUR_DOMAIN/v1/chat/completions", {
  method: "POST",
  headers: {
    "Authorization": "Bearer sk-educa-YOUR_API_KEY",
    "Content-Type": "application/json"
  },
  body: JSON.stringify({
    model: "educa-ai",
    messages: [
      { role: "user", content: "Erstelle einen Wochenplan für Mathe, Klasse 5." }
    ]
  })
});

const data = await response.json();
console.log(data.choices[0].message.content);

using Azure.AI.OpenAI;
using OpenAI.Chat;
using System.ClientModel;

var client = new ChatClient(
    model: "educa-ai",
    credential: new ApiKeyCredential("sk-educa-YOUR_API_KEY"),
    options: new OpenAIClientOptions
    {
        Endpoint = new Uri("YOUR_DOMAIN/v1/")
    }
);

var response = await client.CompleteChatAsync(
    new ChatMessage[]
    {
        new SystemChatMessage("Du bist ein hilfreicher Lehrer."),
        new UserChatMessage("Was ist eine Parabel in Mathe?")
    }
);

Console.WriteLine(response.Value.Content[0].Text);

NuGet: Azure.AI.OpenAI

from langchain_openai import ChatOpenAI

llm = ChatOpenAI(
    model="educa-ai",
    api_key="sk-educa-YOUR_API_KEY",
    base_url="YOUR_DOMAIN/v1/"
)

response = llm.invoke("Erstelle eine Zusammenfassung der Französischen Revolution.")
print(response.content)

Installation: pip install langchain-openai

Sie können Seven Assistant als KI-Assistent in Ihrer IDE nutzen. Konfigurieren Sie in der config.json von Continue.dev:

{
  "models": [
    {
      "title": "Seven Assistant",
      "provider": "openai",
      "model": "educa-ai",
      "apiKey": "sk-educa-YOUR_API_KEY",
      "apiBase": "YOUR_DOMAIN/v1/"
    }
  ]
}