Online Inference API Endpoint Conversation API

Request URL

POST https://vanchin.streamlake.ai/api/gateway/v1/endpoints/chat/completions

Request Header

Authorization: Bearer YOUR_API_KEY
Content-Type: application/json

Request Parameters

Message Format

{
    "messages": [
        {
            "role": "",
            "content": ""
        }
    ],
    "model": "",
    "stream": false,
}

Response Format

For all successfully processed requests, HTTP status code 200 will be returned.

For API call errors, HTTP status codes 4xx or 5xx will be returned as exception responses.

Successful Response (Non-Streaming)

{
    "id": "req-71z2sj-1755282699618223575", // Request ID, uniquely identifies a single request
    "object": "chat.completion", // Object type, indicates this is a chat completion response
    "created": 1741569952, // Unix timestamp (in seconds) when this completion was created
    "model": "", // Model name used to generate this completion
    "choices": [ // A list of chat completion candidates generated by the model (may contain multiple candidate replies)
        {
            "index": 0, // Index of this candidate in the list (0-based)
            "message": { // Message content
                "role": "assistant", // Role type for the generated message content ("assistant")
                "content": "Hello, how can I help you?", // Actual model response content
                "refusal": null, // Reason for refusal (if the AI declined to answer)
                "annotations": [] // Metadata (additional metadata)
            },
            "logprobs": null, // Log probability information (for generation probability analysis)
          	// Reason for generation termination
            "finish_reason": "stop" // Reason for completion（e.g., stop/length/content_filter）
        }
    ],
    "usage": { // Token usage statistics
        "prompt_tokens": 19, // Number of input prompt tokens consumed
        "completion_tokens": 10, // Number of output tokens generated
        "total_tokens": 29, // Total tokens (prompt + completion)
        "prompt_tokens_details": { // Details of input tokens
            "cached_tokens": 0, // Cached tokens count
            "audio_tokens": 0 // Audio tokens count
        },
        "completion_tokens_details": { // Details of output tokens
            "reasoning_tokens": 0, // Tokens used during inference
            "audio_tokens": 0, // Number of audio output tokens
            "accepted_prediction_tokens": 0, // Number of accepted predicted tokens
            "rejected_prediction_tokens": 0 // Number of rejected predicted tokens
        }
    },
    "service_tier": "default" // Service tier (default/scale/etc.)
}

Successful Response (Streaming)

​data: {
    "choices": [
        {
            "delta": {
                "content": "Hello"
            },
            "finish_reason": null,
            "index": 0,
            "logprobs": null
        }
    ],
    "created": 1755282700,
    "event": "model_thought",
    "id": "req-71z2sj-1755282699618223575",
    "model": "qwen3-30b-a3b",
    "object": "chat.completion.chunk",
    "system_fingerprint": null,
    "usage": null
}

......

​data: {
    "choices": [],
    "created": 1755283049,
    "id": "req-pg48yq-1755283048307453384",
    "model": "qwen3-30b-a3b",
    "object": "chat.completion.chunk",
    "system_fingerprint": null,
    "usage": {
        "completion_tokens": 1470,
        "completion_tokens_details": {
            "reasoning_tokens": 454
        },
        "prompt_tokens": 11,
        "total_tokens": 1481
    }
}

data: [DONE]

Error Response

{
    "ResponseMeta": {

        "RequestId": "",
        "ErrorCode": "",
        "ErrorMessage": ""
    }
}