Recipe Generator

Overview

Recipe Generator passes the `name` parameter to a Gemini language model as "Generate a recipe for: " followed by your text, with instructions to be specific about measurements and steps and to answer in English whatever the input language. The model must return JSON that fits a fixed schema, which the API copies field by field into the response.

`name` is the dish title the model chose, which may be tidier or more specific than your input. `description` is a short summary of the dish. `ingredients` is an array of plain strings, each combining quantity, unit and item, such as "1 red bell pepper, sliced"; quantities are not split into separate numeric fields. `instructions` is a single string of steps, numbered and separated by line breaks in typical output. `prep_time` and `cook_time` are free-text durations such as "15 minutes", not numbers. `servings` is an integer.

The model runs at temperature 0.7 with top-p and top-k sampling, so repeated calls for the same dish produce different recipes. Output is capped at 1,000 tokens, which suits everyday dishes; very elaborate requests get a condensed recipe. Units are whatever the model chooses for the dish and are not converted or checked.

Input that fails the text safety check is rejected with a 400 before the model is called, and an empty model reply returns a 500. The recipe is generated text, not a tested recipe: cooking times, temperatures and allergen information should be reviewed before you publish them. There are no plan-dependent fields.

Live Test Recipe Generator API →

Endpoint

One host, one path per API. The block below shows this call in four languages; every one of them is the same HTTP request. Making requests covers the timeouts, retries and parameter rules that apply to all of them. The SDKs wrap the same call in a typed client.

1
2

curl "https://api.apiverve.com/v1/recipe?name=chicken%20fajitas" \
  -H "x-api-key: your_api_key_here"

1
2
3
4
5
6
7
8

const res = await fetch('https://api.apiverve.com/v1/recipe?name=chicken%20fajitas', {
  headers: { 'x-api-key': 'your_api_key_here' },
});

if (!res.ok) throw new Error(`${res.status} ${await res.text()}`);

const { data } = await res.json();
console.log(data);

1
2
3
4
5
6
7
8
9
10

import requests

res = requests.get(
    "https://api.apiverve.com/v1/recipe?name=chicken%20fajitas",
    headers={"x-api-key": "your_api_key_here"},
    timeout=15,
)
res.raise_for_status()

print(res.json()["data"])

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

package main

import (
	"fmt"
	"io"
	"net/http"
)

func main() {
	req, _ := http.NewRequest("GET", "https://api.apiverve.com/v1/recipe?name=chicken%20fajitas", nil)
	req.Header.Set("x-api-key", "your_api_key_here")

	res, err := http.DefaultClient.Do(req)
	if err != nil {
		panic(err)
	}
	defer res.Body.Close()

	out, _ := io.ReadAll(res.Body)
	fmt.Println(string(out))
}

Replace your_api_key_here with the key from your dashboard. When the inputs arrive as a list rather than one at a time, batch requests run up to 200 of them through this same API in a single call.

Authentication

Send your key in the x-api-key header. That is the only auth step — there is no token exchange and no per-endpoint scope to configure. Authentication covers creating, rotating and revoking keys.

Parameters

Sent in the query string. Premium parameters are accepted on every plan but only take effect on plans that include them.

Response

Every API returns the same three top-level keys, so one response handler covers your whole integration: status, error and data. Only data changes shape. Response format covers the envelope, the other output formats and how premium fields are withheld.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

{
  "status": "ok",
  "error": null,
  "data": {
    "name": "Chicken Fajitas",
    "description": "A flavorful and easy-to-make dish with marinated chicken, bell peppers, and onions, served with warm tortillas.",
    "ingredients": [
      "1.5 pounds boneless, skinless chicken breasts, sliced",
      "1 red bell pepper, sliced",
      "1 green bell pepper, sliced",
      "1 large onion, sliced",
      "2 tablespoons olive oil",
      "1 packet (1 ounce) fajita seasoning",
      "1/2 cup water",
      "12 flour tortillas",
      "Optional toppings: sour cream, guacamole, salsa, shredded cheese"
    ],
    "instructions": "1. In a bowl, combine chicken with fajita seasoning and water. Marinate for at least 15 minutes.\n2. Heat olive oil in a large skillet or cast-iron pan over medium-high heat.\n3. Add the chicken and coo…",
    "prep_time": "15 minutes",
    "cook_time": "20 minutes",
    "servings": 4
  }
}

Response fields

Paths are relative to data. Premium fields are absent rather than zeroed on plans that do not include them, so check for presence instead of comparing to 0.

Errors

Read the HTTP status first, then error for the specific reason. The body names the parameter that has to change. Error handling covers the full status list and which of them are worth retrying.

Use cases

Meal Planning Calendars

Meal planning platforms generate weekly dinner schedules by passing suggested dish names and populating shopping lists with the returned ingredient quantities.

Smart Kitchen Displays

When a cook selects a dish on a connected appliance, the interface displays ordered preparation instructions paired with active prep and cook timers.

Grocery Delivery Carts

Supermarket apps turn customer meal requests into full ingredient lists with quantities, adding every required item directly to the digital shopping basket.

Fitness Calorie Trackers

To help users log homemade meals accurately, nutrition logs look up dish concepts to retrieve exact ingredients and estimated serving sizes.

Other ways to use Recipe Generator

Set up Recipe Generator on APIVerve, or reach the same source a different way. Your APIVerve account and credits work on all of them — one key, one balance.

Related

More in Food: