API Reference

Luxeno exposes two OpenAI-compatible endpoints and one Anthropic-compatible endpoint. All endpoints accept standard JSON request bodies and return standard JSON responses.

Authentication

Every request must include your Luxeno API key. Two header formats are accepted — use whichever your client library supports.

Bearer token (recommended)

curl https://api.luxeno.ai/v1/chat/completions \
  -H "Authorization: Bearer sk-silk-xxxxxxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{"model":"luxeno/glm-4-flash","messages":[{"role":"user","content":"Hello"}]}'

x-api-key header

curl https://api.luxeno.ai/v1/chat/completions \
  -H "x-api-key: sk-silk-xxxxxxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{"model":"luxeno/glm-4-flash","messages":[{"role":"user","content":"Hello"}]}'

Get your API key

Go to /keys in your dashboard to create and manage API keys. Keys begin with /keys sk-silk-

Models & Pricing

All prices are in USD per 1 million tokens. Input and output tokens are billed separately where rates differ.

OpenAI-compatible endpoint

Standard Models

Model	Input	Output	Endpoint
luxeno/glm-4-flash	$0.30/1M	$0.30/1M	/v1/chat/completions
luxeno/glm-4	$0.50/1M	$0.50/1M	/v1/chat/completions

Standard models charge the same rate for all input token types, including cached tokens.

Cache-Aware Models

Model	Input	Output	Cached Read	Cache Write	Cache Storage	Endpoint
luxeno/claude-sonnet	$3.00/1M	$15.00/1M	$0.30/1M	$3.75/1M	—	/v1/chat/completions

Anthropic-compatible endpoint

Standard Models

Model	Input	Output	Endpoint
claude-sonnet-4-20250514	$0.30/1M	$0.30/1M	/v1/messages
glm-4-flash	$0.30/1M	$0.30/1M	/v1/messages
glm-4.7	$0.50/1M	$0.50/1M	/v1/messages
glm-5	$0.50/1M	$0.50/1M	/v1/messages

Standard models charge the same rate for all input token types, including cached tokens.

claude-sonnet-4-20250514, glm-4-flashclaude-sonnet-4-20250514, glm-4-flash, and other Anthropic-format models are routed through Zhipu's Anthropic-compatible endpoint at the same $0.30/1M rate — significantly cheaper than native Anthropic pricing.

POST /v1/chat/completions

OpenAI-compatible chat endpoint. Accepts the same request schema as openai.chat.completions.create() — drop-in replacement by changing the base URL and model name.

Request

POST /v1/chat/completions

POST /v1/chat/completions
Content-Type: application/json
Authorization: Bearer sk-silk-xxxxxxxxxxxx

{
  "model": "luxeno/glm-4-flash",
  "messages": [
    { "role": "system", "content": "You are a helpful assistant." },
    { "role": "user",   "content": "Explain quantum entanglement in one sentence." }
  ],
  "temperature": 0.7,
  "max_tokens": 256
}

Response

200 OK

{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "created": 1748000000,
  "model": "luxeno/glm-4-flash",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Quantum entanglement is a phenomenon where two particles become correlated such that measuring one instantly determines the state of the other, regardless of distance."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 31,
    "completion_tokens": 38,
    "total_tokens": 69
  }
}

Streaming (SSE)

Set "stream": true to receive server-sent events. The final chunk before [DONE] includes the usage object for billing transparency.

Streaming example

POST /v1/chat/completions
Content-Type: application/json
Authorization: Bearer sk-silk-xxxxxxxxxxxx

{
  "model": "luxeno/glm-4-flash",
  "messages": [{ "role": "user", "content": "Count to five." }],
  "stream": true
}

--- SSE response ---
data: {"id":"chatcmpl-xyz","object":"chat.completion.chunk","choices":[{"delta":{"role":"assistant","content":"1"},"index":0}]}

data: {"id":"chatcmpl-xyz","object":"chat.completion.chunk","choices":[{"delta":{"content":", 2"},"index":0}]}

data: {"id":"chatcmpl-xyz","object":"chat.completion.chunk","choices":[{"delta":{},"finish_reason":"stop","index":0}],"usage":{"prompt_tokens":12,"completion_tokens":10}}

data: [DONE]

POST /v1/messages

Anthropic Messages API-compatible endpoint. Accepts the same request schema as the official Anthropic SDK. Use this endpoint with tools that call the Anthropic API directly (e.g. Claude Code, Cursor in Anthropic mode).

anthropic-version header

Include anthropic-version: 2023-06-01 in your request. If omitted, the gateway defaults to that version automatically.

Request

POST /v1/messages

POST /v1/messages
Content-Type: application/json
Authorization: Bearer sk-silk-xxxxxxxxxxxx
anthropic-version: 2023-06-01

{
  "model": "claude-sonnet-4-20250514",
  "max_tokens": 1024,
  "system": "You are a concise technical writer.",
  "messages": [
    { "role": "user", "content": "What is a B-tree index?" }
  ]
}

Response

200 OK

{
  "id": "msg_01AbCdEf",
  "type": "message",
  "role": "assistant",
  "content": [
    {
      "type": "text",
      "text": "A B-tree index is a balanced tree data structure that maintains sorted data and allows searches, insertions, deletions, and range queries in O(log n) time."
    }
  ],
  "model": "claude-sonnet-4-20250514",
  "stop_reason": "end_turn",
  "usage": {
    "input_tokens": 24,
    "output_tokens": 42
  }
}

Streaming SSE event types

Anthropic SSE uses named event types rather than a single data: stream. Key events:

message_start — contains input token count in message.usage.input_tokens
content_block_delta — carries text chunks in delta.text
message_delta — contains output token count in usage.output_tokens
message_stop — signals end of stream

Anthropic SSE stream

data: {"type":"message_start","message":{"id":"msg_01AbCd","type":"message","role":"assistant","usage":{"input_tokens":24,"output_tokens":0}}}

data: {"type":"content_block_start","index":0,"content_block":{"type":"text","text":""}}

data: {"type":"content_block_delta","index":0,"delta":{"type":"text_delta","text":"A B-tree index"}}

data: {"type":"content_block_delta","index":0,"delta":{"type":"text_delta","text":" is a balanced tree..."}}

data: {"type":"message_delta","delta":{"stop_reason":"end_turn"},"usage":{"output_tokens":42}}

data: {"type":"message_stop"}

Rate Limits

Each user account is limited to 60 requests per minute across both the /v1/chat/completions and /v1/messages endpoints combined (sliding window).

429 response

HTTP/1.1 429 Too Many Requests
Retry-After: 12
Content-Type: application/json

{
  "error": {
    "message": "Rate limit exceeded",
    "type": "rate_limit_error",
    "code": 429
  }
}

Retry-After

Always read the Retry-After header (value in seconds) before retrying. Immediate retries will be rejected and count against your limit.

Error Codes

All errors follow OpenAI-format envelopes: { error: { message, type, code } }. Anthropic-format errors on the /v1/messages endpoint use { type: 'error', error: { type, message } }.

Status	Type	Cause	Resolution
400	invalid_request_error	Bad model name or malformed body	Check the model name and request schema
401	authentication_error	Missing or invalid API key	Verify your key in the dashboard
402	payment_required	Account balance is zero	Top up your balance at /billing
429	rate_limit_error	60 RPM limit exceeded	Respect the Retry-After header
500	api_error	Unexpected server error	Retry with exponential back-off
503	api_error	Upstream gateway unavailable	Retry; check status page

Playground

Try the API directly in your browser. The playground calls /api/v1/chat/completions using your first active API key. Requests are billed to your account at normal rates.

Loading playground...