Start building with Neiroport

Your first SDK call

Examples use an identifier from the current catalogue. Choose the available variant you need.

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.NEIROPORT_API_KEY,
  baseURL: "https://neiroport.com/v1"
});

const response = await client.chat.completions.create({
  model: "qwen3.6-27b",
  messages: [{ role: "user", content: "Hello!" }],
  max_tokens: 1024
});

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

Three API formats

Choose a request format supported by the model. All supported protocols use your Neiroport key and the service API URL.

curl https://neiroport.com/v1/messages \
  -H "x-api-key: $NEIROPORT_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "claude-fable-5",
  "max_tokens": 1024,
  "messages": [
    {
      "role": "user",
      "content": "Hello!"
    }
  ]
}'

curl "https://neiroport.com/v1beta/models/gemini-2.5-flash:generateContent" \
  -H "x-goog-api-key: $NEIROPORT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "contents": [
    {
      "role": "user",
      "parts": [
        {
          "text": "Hello!"
        }
      ]
    }
  ],
  "generationConfig": {
    "maxOutputTokens": 1024
  }
}'

Model catalogue and variants

The catalogue updates automatically. Identifiers and suffixes are preserved: variants have distinct model IDs and tariffs. Your preference determines the route.

curl https://neiroport.com/v1/models \
  -H "Authorization: Bearer $NEIROPORT_API_KEY"

Routing preferences

Compare offers in Model suppliers and pin a supplier separately for each model and client key. Displayed prices include markup. Pins take priority over automatic routing and apply in SDK calls without extra parameters.

Pinned routes do not fall back to other suppliers. If a tariff changes or a supplier becomes unavailable, the request is rejected before dispatch; select a supplier again or restore automatic routing. Select the same client key in the playground to test its route.

Choose a preference when creating an API key; you can change it later without replacing the key. All calls using this key follow the selected preference. The playground has a separate preference selector.

Prices and availability may differ by preference. The catalogue shows the selected preference’s tariff. Global and individual markup apply across all preferences; unconfirmed tariffs cannot be used for generation. The guaranteed preference uses only guaranteed routes.

Model details show API formats, billing type, input, output, cache read and write, or a per-request price. Reference prices are not active tariffs until confirmed.

Endpoint reference

Relative paths use the /v1 Base URL. Gemini /v1beta paths start at the service root. Methods marked “Unavailable” are not currently connected in Neiroport.

curl https://neiroport.com/v1/responses \
  -H "Authorization: Bearer $NEIROPORT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "qwen3.6-27b",
  "input": "Hello!",
  "max_output_tokens": 1024,
  "store": false
}'

curl https://neiroport.com/v1/embeddings \
  -H "Authorization: Bearer $NEIROPORT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "gemini-embedding-2-preview",
  "input": "Text to embed"
}'

curl https://neiroport.com/v1/images/generations \
  -H "Authorization: Bearer $NEIROPORT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "google-imagen-4",
  "prompt": "A quiet mountain lake",
  "n": 1,
  "size": "1024x1024"
}'

Troubleshooting

Use error.code and request_id to diagnose a failure. API errors and an extended compatibility reference are listed separately. Internal routing details are not exposed to clients.

{
  "error": {
    "code": "model_not_found",
    "message": "model_not_found"
  },
  "request_id": "http_xxxxxxxxxxxx"
}

Billing and streaming usage

stream = client.chat.completions.create(
    model="qwen3.6-27b",
    messages=[{"role": "user", "content": "Hello!"}],
    max_tokens=1024,
    stream=True,
    stream_options={"include_usage": True}
)

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

Send a unique Idempotency-Key with paid POST requests. Reusing it returns 409 and the original request_id. Check usage history before retrying an interrupted stream.