Developer API

Every model, one API key.

OpenAI-compatible endpoints for chat, image and video generation across all the models in your plan. Points-based pricing, automatic failover, no provider accounts to manage.

Get your API key See plans

One key. Every model.

The AI Fiesta API is OpenAI-compatible. Point any OpenAI SDK at our base URL, use your AI Fiesta key, and pick a model from the table below. Requests are metered in points from your balance exactly like the app — failed calls are never charged, and provider keys never leave our servers.

Base URL

https://aifiesta-api.dev.codexcape.solutions/v1

Auth header

Authorization: Bearer af_YOUR_API_KEY

Rate limit

60 requests / minute per key

Quick start

Chat completion with curl and with the OpenAI SDKs.

bash

curl https://aifiesta-api.dev.codexcape.solutions/v1/chat/completions \
  -H "Authorization: Bearer af_YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-sonnet-5",
    "messages": [{"role": "user", "content": "Say hello in five words."}]
  }'

python

from openai import OpenAI

client = OpenAI(base_url="https://aifiesta-api.dev.codexcape.solutions/v1", api_key="af_YOUR_API_KEY")

resp = client.chat.completions.create(
    model="claude-sonnet-5",
    messages=[{"role": "user", "content": "Say hello in five words."}],
    stream=True,
)
for chunk in resp:
    print(chunk.choices[0].delta.content or "", end="")

javascript

import OpenAI from "openai";

const client = new OpenAI({ baseURL: "https://aifiesta-api.dev.codexcape.solutions/v1", apiKey: "af_YOUR_API_KEY" });

const res = await client.chat.completions.create({
  model: "claude-sonnet-5",
  messages: [{ role: "user", content: "Say hello in five words." }],
});
console.log(res.choices[0].message.content);
console.log(res.aifiesta); // { points_charged, points_balance, upstream_model }

What comes back

Standard OpenAI chat.completion objects, plus an aifiesta block with the points charged for the call and your remaining balance. With stream: true you get chat.completion.chunk frames and a final data: [DONE].

Send images as image_url content parts (https or data URLs). Set max_tokens and temperature as usual; we translate vendor quirks for you.

Endpoints

All paths are relative to the base URL and require your key.

GET/modelslive catalog

Returns every model available to you with type (chat / image / video), points_per_request and context window.

POST/chat/completions

OpenAI chat schema. Body: model, messages, optional stream, max_tokens, temperature. Cost: the model's points, charged only on success.

POST/images/generations

Synchronous. Body: model, prompt, optional n (1–4), aspect_ratio (e.g. 16:9) or OpenAI size, optional image_url reference for edits. Returns hosted URLs. Failed images are refunded.

bash

curl https://aifiesta-api.dev.codexcape.solutions/v1/images/generations \
  -H "Authorization: Bearer af_YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "gemini-flash-image", "prompt": "a paper boat on a puddle after rain", "aspect_ratio": "16:9"}'

POST/videos/generationsasync

Returns 202 with a job id; poll GET /videos/generations/{id} every few seconds until status is completed (then url is set) or failed (points refunded). Body: model, prompt, aspect_ratio, duration seconds, optional image_url first frame.

bash

curl https://aifiesta-api.dev.codexcape.solutions/v1/videos/generations \
  -H "Authorization: Bearer af_YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "veo-3-1", "prompt": "slow dolly through a misty pine forest at dawn", "aspect_ratio": "16:9", "duration": 6}'

# then
curl https://aifiesta-api.dev.codexcape.solutions/v1/videos/generations/JOB_ID -H "Authorization: Bearer af_YOUR_API_KEY"

GET/usage

Requests and points spent by this key, your current balance and the most recent calls.

Models & prices

Live from the catalog — what you see here is what the API accepts as `model`.

Errors & limits

401 invalid_api_key — missing, wrong or revoked key.
402 insufficient_points — top up or redeem a code, then retry.
403 model_not_allowed / api_key_disabled / api_access_disabled — an administrator has restricted this model, key or account; contact support.
404 not_found — unknown model; check GET /models.
429 rate_limited — 60 requests per minute per key; back off for details.retry_after seconds.
429 quota_exceeded — an admin-set daily request or monthly points cap for your account was reached; GET /usage shows the limits under quota.
502 provider_error / 503 model_unavailable — upstream trouble after our retries and fallbacks; nothing was charged.

Error bodies are OpenAI-shaped: {"error": {"message", "type", "code"}}.

Need something the API does not cover yet? Contact support from your account.