whatevr API

Log food, exercise, water and weight — or read back a day's nutrition — from any app, script, or AI assistant.

Base URL https://api.trywhatevr.app

Two ways in

Connect an AI assistant

Hook up Claude, ChatGPT, Grok or Gemini and log by chat. No code. Setup guide →

Build with the REST API

A token, a base URL, and nine endpoints. For scripts, hardware, or your own app. Quick start →

Quick start

Make your first call

  1. Open the whatevr app → Settings → Assistants & API → create a token. It looks like wtvr_live_… and is shown once.
  2. Send it as a bearer token on every request. Pick a scope: read, log, or full.
# log a meal in plain language
curl https://api.trywhatevr.app/v1/ai/log \
  -H "Authorization: Bearer wtvr_live_…" \
  -H "Content-Type: application/json" \
  -d '{"text":"two eggs and black coffee"}'

Reference

Endpoints

All under /v1. Full schema in the OpenAPI spec.

POST/v1/ai/logLog anything in natural language
POST/v1/logs/exerciseLog structured exercise
POST/v1/logs/weightLog a weigh-in
POST/v1/waterLog water
POST/v1/logs/updateCorrect an existing food log
GET/v1/todayToday's totals, targets, streak, feed
GET/v1/dayA specific day's data
GET/v1/progressDaily series + weight trend
GET/v1/meUser, profile, targets, metrics

For AI agents

MCP endpoint

whatevr is also a Model Context Protocol server, so agents can log and read on your behalf. Point any MCP client at:

https://api.trywhatevr.app/mcp   # bearer token in the header
https://api.trywhatevr.app/mcp/<token>   # or the token in the URL

Most people don't touch this directly — the connect guide walks you through adding it to Claude, ChatGPT, Grok or Gemini.

Good to know

Auth
Authorization: Bearer wtvr_live_…
Scopes
read · log · full
Timestamps
Local, no zone. Pass tz_offset_min so days land right.
Rate limits
~100 req/min · ~15/min on /v1/ai/log · 429 + Retry-After
Copied