AislePrompt Features — AI Meal Planning, Pantry & Instacart

What AislePrompt does for you in one screen: describe a meal in plain English and an AI builds the recipe + grocery cart; browse a 91,700+ recipe catalog filtered by diet, time, and ingredients you already own; check out in one tap on Instacart at Target, Walmart, Whole Foods, Costco, and most US grocers. Bilingual, family-shareable, and free — no account required. Each feature below is a one-paragraph overview followed by linked deep-dives. Developers and integrators looking for data boundaries and edge cases will find the technical specifics inline in each section.

Try AislePrompt free → Shop on Instacart →

AI meal-planning chat

Describe a craving, dietary constraint, or ingredient list and the AI picks recipes and builds a cart — bilingual (English/Spanish), no account required. Refine mid-session: "make it gluten-free," "swap the salmon," "scale to 4 servings" — the suggestion set updates in place without restarting. For multi-turn examples and accessibility details, see Chat; for pricing, see the FAQ; for device compatibility and the full accessibility posture, see the Accessibility & language support section on this page.

Technical details for developers ▸
Context model: each turn appends the parsed constraint to a running recipe-filter state, so new instructions refine — not restart — the suggestion set. The input parser accepts meal descriptions, dietary flags (keto, gluten-free, vegan, halal), allergen exclusions, serving-size multipliers, and ingredient-override tokens. Constraints stack and are evaluated against the full catalog on each turn; the session resets only when the user explicitly clears it. Output is a structured ingredient list ready for the shopping-list stage, not a prose response. Token limits: 8 turns of history are retained per session; older turns are summarised into a compact constraint preamble so the parser keeps state without bloating prompt length. Failure mode: when no recipe in the catalog satisfies all stacked constraints, the model is invoked to synthesise a new recipe row that meets them, which is then persisted with the same schema as catalog rows so downstream ingredient deduplication and Instacart matching are identical.

Recipe catalog

Browse 91,700+ recipes filtered by cuisine, dietary tag, cook time, or ingredient. Results rank by relevance weighted by community rating so the most widely-liked recipes surface first. Every recipe links directly to Instacart at the ingredient level and to Amazon for related kitchen items. Organized navigation: Latest 500 Recipes (date-sorted); Curated Recipe Shortlists (single-intent roundups); Full Filterable Recipe Directory (full catalog with tag/cuisine/time filters).

Technical details for developers ▸
Data model: each recipe row carries rating (1–5 float), vote_count, macros (calories, protein_g, carbs_g, fat_g as integers), cook_time_min, prep_time_min, and a dietary_tags set (vegan, vegetarian, gluten-free, keto, halal, dairy-free, nut-free — each independently filterable). Search ranks candidates by semantic distance weighted by rating × log(vote_count + 1); dietary-tag filtering is applied before the ranking pass so only compliant recipes enter the score queue.

Pantry tracker

A persistent ingredient inventory that integrates with the AI at three points: (1) at list-build time, owned quantities are subtracted from required quantities before the Instacart handoff — if the pantry holds 2 cups of rice and a recipe requires 1 cup, rice is dropped from the cart; (2) items with an expiry date within 3 days receive a score boost applied before the semantic ranking pass, so near-expiry ingredients surface at the top of recipe suggestions; (3) the weekly meal plan calendar deducts pantry-owned items from the consolidated weekly cart before sending to Instacart. Items enter the pantry via barcode scan, photo of a shelf (vision model), or receipt OCR — the full scanning pipeline is on the AI Pantry Scanner page. The user-facing workflow is on Pantry.

Shopping list

Ingredients from chat, recipes, and the meal plan all flow into a single shared list. AislePrompt deduplicates across recipes (two recipes calling for "1 lemon" each produce "2 lemons"), normalizes units (oz ↔ cups ↔ grams), and groups items by Instacart aisle (Produce → Dairy → Meat → Pantry → Frozen) so the list maps to how you actually shop. Brand-aware UPC matching remembers the brands you previously ordered on Instacart, pre-selecting the same SKU next time. Full details and guest-mode info on the Shopping List page.

Meal plan calendar

A 7-day drag-and-drop calendar. At cart-build time, ingredients across all recipe slots are merged, duplicates summed, units normalized, and the result is sent to Instacart as a single consolidated weekly cart. Daily and weekly calorie/macro totals are computed from per-recipe nutrition rows and update in real time as slots change. Slot behavior: locked slots are excluded from AI re-plan suggestions; leftover slots suppress their ingredient set from the cart calculation. Pantry deduction runs last — pantry-owned quantities are subtracted from the merged weekly cart before the Instacart handoff. Walkthrough on Meal Plan.

Kitchen + grocery shop

A 62,600+ item catalog across 14 categories (cookware, knives, appliances, bakeware, pantry staples, and more). Each product entry carries an Amazon ASIN and, where stocked, an Instacart product ID, enabling side-by-side price compare from both retail channels. Product pages include editorial notes, linked specs, and direct add-to-cart buttons for both retailers. Browse on Kitchen Shop.

Family sharing

Up to 8 household members share one pantry inventory, one shopping list, and one meal plan calendar, all synced in real time across every member's devices. Per-member dietary profiles (keto, gluten-free, nut-free, etc.) keep the AI from suggesting meals that conflict with any individual's restriction. Role system: Owner can invite/remove members; Members can edit shared state; Viewers can read-only. Conflict resolution (two members editing the same list slot simultaneously) is last-write-wins with a 3-second merge window. Full invite flow, privacy controls, and role reference on Family Sharing.

Accessibility & language support

AislePrompt is a mobile-first web app tested against iOS Safari and Android Chrome. The interface is bilingual (English + Spanish) — the full UI, AI chat, and translated recipe titles and instructions switch language from the header. Screen reader support: semantic HTML5 landmarks, ARIA labels on interactive controls, and focus-visible outlines throughout. The app uses system fonts and respects OS-level text-size preferences; there is no custom font-size override that could conflict with browser zoom. High-contrast mode follows the OS preference via the CSS prefers-contrast media query on interactive elements. Guest mode requires no login and no cookie consent beyond functional cookies — low friction for assistive-technology users. If you encounter an accessibility barrier, please send feedback.

FAQ

Is AislePrompt free?

Yes — free, no credit card required. We earn a small affiliate fee when you check out on Instacart or Amazon at no cost to you.

Does AislePrompt work on iPhone and Android?

Yes. AislePrompt runs in any modern browser — iOS Safari, Android Chrome, desktop browsers. PWA install supported.

Which retailers does Instacart support?

Instacart covers most US grocery stores: Target, Walmart, Whole Foods, Costco, Wegmans, Publix, Aldi, and many regional chains.

Is AislePrompt accessible for screen reader users?

AislePrompt uses semantic HTML5 landmarks, ARIA labels on interactive controls (add-to-cart, pantry scan, recipe swap), and focus-visible outlines so keyboard-only and screen-reader navigation works throughout the app. The full interface is bilingual (English + Spanish). If you hit a barrier, use the feedback form to report it.