grammY TestingProduction-grade testing for Telegram bots

Drive your real bot in-process. Capture every API call. Assert on replies — no tokens, no network, no waiting.

⚡

In-Process Testing

Your bot runs entirely in-process. No Telegram token, no network round-trips, no flaky timeouts. Tests are deterministic and fast.

🎭

Real Bot, Synthetic Updates

Uses your actual grammY bot with a transformer that intercepts outgoing calls. No mocking of grammY internals — every handler runs exactly as in production.

🏗️

Two-Layer API

A high-level actor API (User, Group, Channel) for readable tests, plus a low-level escape hatch for edge cases. Use what fits each test.

🧪

Any Test Framework

Works with Vitest, Jest, and Deno's built-in test runner. Zero framework-specific dependencies in the library itself.

Quick look

import { prepareBot } from 'grammy-testing';
import { Bot } from 'grammy';
import { describe, expect, it } from 'vitest';

const bot = new Bot('token');
bot.command('start', (ctx) => ctx.reply('Hello!'));

it('replies to /start', async () => {
  const { chats } = await prepareBot(bot);
  const user = chats.newUser();

  await user.sendCommand('/start');

  expect(user.replies.lastOrThrow().text).toBe('Hello!');
});

Install

npmyarnpnpmDeno

npm install --save-dev grammy-testing

yarn add --dev grammy-testing

pnpm add --save-dev grammy-testing

import { prepareBot } from 'npm:grammy-testing';