Endpoints

Responses

Use POST /responses on api14 with the OpenAI Responses API format, including input items, tools, streaming and structured output.

Try it in the playground

The Responses endpoint accepts the OpenAI Responses API format, so client.responses.create(...) in the official SDKs works without changes. It runs on the same models as Chat completions. api14 forwards the body upstream unchanged and bills the usage that comes back.

POST https://api14.com/api/v1/responses
Authorization: Bearer sk-api14-...
Content-Type: application/json

Request

Parameter	Type	Description
`model`	string, required	Model id, e.g. `openai/gpt-5-mini`.
`input`	string \| array, required	A prompt string or an array of input items (`{ role, content }` messages, `function_call_output`, …).
`instructions`	string	System-level instructions.
`stream`	boolean	Stream semantic events (`response.output_text.delta`, `response.completed`, …) over SSE.
`max_output_tokens`	integer	Upper bound on generated tokens.
`temperature`, `top_p`	number	Sampling controls.
`tools`, `tool_choice`, `parallel_tool_calls`		Function tools. Web search is also available as a tool or through `plugins`.
`text`	object	Output format, e.g. `{ "format": { "type": "json_schema", "name": "...", "schema": {...} } }`.
`reasoning`	object	`{ "effort": "low" \| "medium" \| "high" }`. See Reasoning.
`provider`, `models`, `plugins`		Same gateway extensions as chat completions.

The gateway doesn't store your conversations, so pass the full context in input with every call. Stateful options like previous_response_id may not be supported.

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api14.com/api/v1",
  apiKey: process.env.API14_API_KEY,
});

const result = await client.chat.completions.create({
  "model": "openai/gpt-5-mini",
  "messages": [
    {
      "role": "user",
      "content": "Explain vector databases in two sentences."
    }
  ]
});
console.log(result.choices[0].message.content);

The example above uses chat completions. With the Node SDK, the Responses equivalent looks like this:

const response = await client.responses.create({
  model: "openai/gpt-5-mini",
  input: "Write a haiku about APIs.",
});
console.log(response.output_text);

Response

{
  "id": "resp-1760000000-abc123",
  "object": "response",
  "created_at": 1760000000,
  "model": "openai/gpt-5-mini",
  "status": "completed",
  "output": [
    {
      "type": "message",
      "role": "assistant",
      "content": [{ "type": "output_text", "text": "Requests in flight,\nJSON answers drifting back,\none key opens all." }]
    }
  ],
  "usage": {
    "input_tokens": 14,
    "output_tokens": 22,
    "total_tokens": 36,
    "input_tokens_details": { "cached_tokens": 0 },
    "output_tokens_details": { "reasoning_tokens": 0 },
    "cost": 0.0000475
  }
}

The gateway reads usage from the final object, or from the response.completed event when streaming, and debits cost in USD from your balance.

Errors

Same as chat completions: 400 model_required, 400 invalid_json, 401 invalid_api_key, 402 insufficient_balance, 403 model_not_allowed / plan_required, 404 model_not_found, 413 payload_too_large, 429 rate_limited / daily_limit_reached, 502 upstream_unavailable. See Errors.