ai service

ai helpers that plug into any stack. Browse the catalog below and pick the helpers that fit your workflow.

Grab your Dataczar API key

Every helper runs with the same simple API key. Sign in (or create a free Dataczar account) at Dataczar Connect to issue keys in seconds, then drop the key into any `ak=` parameter and start shipping.

Get your API key Browse helpers

Tip: reuse the same key across these helpers, or rotate keys per project from the Connect dashboard.

GETPOSTApply AI-driven edits to an image.

Uploads or references an existing image plus an edit prompt (and optional mask) to let the configured provider redraw specific regions. Works for foreground swaps, cleanup, background replacements, etc.

GETList the available AI helpers and their payloads.

Returns a catalog of the AI entry points (`/ai/text`, `/ai/image`, `/ai/interpret`, `/ai/edit`) along with the expected fields for each helper. This is the canonical way to fetch up-to-date docs straight from the Lambda.

GETPOSTGenerate branded images from a short brief.

Turns a text prompt into a rendered image using the configured AI provider (OpenAI when `AI_PROVIDER=openai`, deterministic mock otherwise). Size and style defaults follow the Lambda env vars (`AI_IMAGE_MODEL`, `AI_IMAGE_SIZE`) but you can override them per request.

GETPOSTDescribe or summarize an image with multimodal AI.

Accepts an image (either by URL or base64 payload) and optional instructions, then runs it through the configured multimodal model. Useful for auto-captioning screenshots, validating mockups, or asking follow-up questions about a visual.

GETPOSTGenerate text with the configured AI provider.

Accepts a prompt and optional model/temperature knobs, then returns raw provider output. Defaults to `gpt-4o-mini` when `AI_PROVIDER=openai`, or a deterministic mock response when the Lambda is running in mock mode.