Stream generate content | EmpirioLabs AI Docs

curl -X POST https://api.empiriolabs.ai/v1beta/models/qwen3-7-max:streamGenerateContent \
     -H "Authorization: Bearer <token>" \
     -H "Content-Type: application/json" \
     -d '{
  "contents": [
    {
      "role": "user",
      "parts": [
        {
          "text": "Can you provide a summary of the latest advancements in renewable energy?"
        }
      ]
    }
  ],
  "systemInstruction": {
    "role": "model",
    "parts": [
      {
        "text": "You are a helpful assistant providing concise and accurate information."
      }
    ]
  },
  "generationConfig": {
    "temperature": 0.5,
    "topP": 0.9,
    "topK": 40,
    "maxOutputTokens": 150,
    "stopSequences": [
      "\n\n"
    ],
    "candidateCount": 1,
    "responseMimeType": "application/json",
    "thinkingConfig": {
      "thinkingBudget": 1000,
      "includeThoughts": false
    }
  },
  "tools": [],
  "toolConfig": {
    "functionCallingConfig": {
      "mode": "AUTO",
      "allowedFunctionNames": []
    }
  },
  "safetySettings": []
}'

{
  "event": "data: {\"candidates\":[{\"content\":{\"parts\":[{\"text\":\"Recent advancements in renewable energy include improvements in solar panel efficiency, the development of offshore wind farms, and breakthroughs in energy storage technologies such as solid-state batteries.\"}]} }],\"finishReason\":null,\"usageMetadata\":null}\n\n"
}

Streaming variant of generateContent. Returns Server-Sent Events when called with ?alt=sse (what the official SDKs use); each event is a partial GenerateContentResponse with the incremental text in candidates[0].content.parts[0].text. The final event carries finishReason and usageMetadata. There is no [DONE] sentinel; the stream simply ends.

Streaming variant of generateContent. Returns Server-Sent Events when called with `?alt=sse` (what the official SDKs use); each event is a partial `GenerateContentResponse` with the incremental text in `candidates[0].content.parts[0].text`. The final event carries `finishReason` and `usageMetadata`. There is no `[DONE]` sentinel; the stream simply ends.

Authentication

AuthorizationBearer

Pass your EmpirioLabs API key as a bearer token. The Anthropic-style x-api-key header is also accepted on every endpoint.

Path parameters

modelstringRequired

An EmpirioLabs chat model id, for example qwen3-7-max.

Query parameters

altenumOptional

Set to sse for Server-Sent Events.

Allowed values:

Request

This endpoint expects an object.

contentslist of objectsRequired

Conversation turns. Roles are user and model.

systemInstructionobjectOptional

generationConfigobjectOptional

toolslist of objectsOptional

toolConfigobjectOptional

safetySettingslist of objectsOptional

Accepted for compatibility.

Response

A stream of partial GenerateContentResponse events (SSE).

Errors

400

Bad Request Error

401

Unauthorized Error

429

Too Many Requests Error