Chatbase

Chatbase

Chatbase Inc. · マーケティング · チャットボット

Chatbase is a no-code chatbot builder for AI customer support. It's an AI chatbot platform that trains agents on your own documents, website, and tickets. You add your content, pick a model, and deploy the agent across a website widget, email, WhatsApp, Slack, Messenger, Instagram, or inbound phone calls. It started as a way for small teams to answer repetitive questions, and it now sells an enterprise tier with SSO, SLAs, and HIPAA support.

Chatbase の画面プレビュー

Chatbase について

What Is Chatbase

Chatbase is an AI agent platform built around customer support. You feed it your help center articles, PDFs, Notion pages, or Zendesk tickets, and the agent answers questions using that material instead of inventing answers. The pitch is simple: cut the queue of "where's my order" and "how do I reset my password" tickets without hiring more staff. No code required.

The product has moved well past the basic embed-a-chat-widget era. Agents can now take real actions inside connected systems, like looking up a Stripe subscription or checking a Shopify order, and hand a conversation to a human with the transcript and an AI summary attached. Phone support works too, through Twilio or your own SIP trunk. Voice is one of the pricier channels to run at scale.

The main limits are practical ones. Training content is capped by plan (1 MB on Free, up to 40 MB on Pro), and support tickets count as a data source only on paid tiers. Message credits, not conversations, are what you're really paying for, and a single user question can burn several credits if the agent runs an action. Prices start at $40/month. So what does that buy you? 700 message credits on Hobby, which is fine for a small site and thin for a busy one. If your documentation is messy or outdated, the agent will faithfully repeat your mess.

Getting Started

  1. Sign up at chatbase.co and create a new AI agent from the dashboard.
  2. Train it by adding sources: upload files (PDF, DOC, DOCX, TXT up to 20 MB each), point it at a website URL, paste text snippets, or paste Q&A pairs. The agent picks up changes within seconds, and there's no retrain button to click.
  3. Test in the Playground, where you can preview the agent across channels and adjust its instructions and tone in real time.
  4. Deploy by enabling a channel: a floating chat bubble, a hosted help page, Slack, WhatsApp, Email, or a phone number. Most website setups are a single script tag.
  5. Monitor conversations in the Activity log and check Analytics to see what people actually ask.

AIツール情報

Chatbaseの料金、対応プラットフォーム、性能を簡単に確認できます。

無料プランはい
有料プラン$40 - $500/mo (plus custom Enterprise pricing)
プラットフォームWeb (browser-based; deploys to websites, messaging apps, email, phone, and via iOS/Android SDKs)
開発元Chatbase Inc.
カテゴリマーケティング · チャットボット
リリース日May 2023
最終更新Sep 2026
サイト訪問数402.2K
サイト世界ランキング93.5K
API提供状況はい

こんな人におすすめ

このツールが最も力を発揮するユーザー、タスク、シーン。

ユーザー

  • Small support teams drowning in repeat questions
  • Ecommerce stores on Shopify or Stripe
  • SaaS companies with a decent help center
  • Agencies and freelancers building chatbots for clients

タスク

  • Answering product and pricing questions
  • Triaging and escalating tickets
  • Booking meetings
  • Collecting leads

シーン

  • Handling overnight support for a small brand
  • Deflecting first-line tickets before they reach Zendesk or Intercom
  • Standing up a beta chatbot in an afternoon

主な機能

No-code agent builder with instant training

You build and train a custom AI agent from a dashboard, no programming required. Add files, a website URL, text snippets, Notion pages, or Q&A pairs and the agent picks them up within seconds. There's no retrain button, which removes a step that trips people up on other platforms. Edit a source and the agent learns the change before you close the tab.

A chatbot for customer service, not a general assistant

Chatbase is aimed at support work rather than open-ended chat. The agent answers from your help center and ticket history, and it stays inside the guardrails you set. That focus is why the escalation and action features feel complete instead of bolted on. It won't write your blog posts. It will handle the same password question for the hundredth time without complaining. Narrow scope, done well.

Multi-channel deployment from one agent

A single agent goes to a website chat bubble, a hosted help page, Slack, WhatsApp, Messenger, Instagram, email, and inbound phone calls. You enable and configure each channel separately, so you can start with web chat and add voice later without rebuilding. Native iOS and Android SDKs exist for teams embedding the agent inside their own apps. One agent, many doors.

Actions that change things, not just answer

Beyond answering questions, agents can run actions in connected systems: look up a Stripe subscription, browse Shopify products and manage carts, create cases in Salesforce, or post to Slack. That's the difference between a chatbot and a support tool. A custom action lets you call your own API or run browser code, which is how teams wire the agent into internal tools the standard integrations don't cover. Plenty of power here.

Human escalation with full context

When a conversation needs a person, the agent hands off with the transcript and an AI-generated summary attached, so the human doesn't start from zero. Rules are written in plain language rather than a code editor. Escalations route to Chatbase's own helpdesk or a connected platform like Zendesk or Salesforce. Fast handoff, no lost context.

Voice and phone support

Agents can take inbound calls through Twilio or your own SIP trunk, and transfer a live call to a human by forwarding to a phone number. Concurrent call handling is plan-dependent, and voice is one of the pricier channels to run at volume. Test it before you commit.

Procedures for multi-step workflows

Procedures let you define a workflow with a trigger, then the agent follows those steps end to end instead of improvising. This matters for anything with a required order, like a refund flow or an onboarding sequence, where a free-form answer would skip a step. Steps stay on rails.

Backstage, an AI copilot for managing the agent

Backstage is an operations layer where you describe changes in plain English and it handles the configuration. Instead of hunting through settings pages, you tell it what you want the agent to do differently. It won't replace understanding the settings, but it speeds up routine tweaks. Think of it as a copilot for admin work.

Security and compliance controls

Chatbase holds SOC 2 Type II and is GDPR compliant, with HIPAA Business Associate Agreements available on Enterprise. Data is encrypted at rest and in transit, and customer data isn't used to train shared models. Enterprise customers can choose data residency across US, EU, or APAC regions. For regulated teams, that matters more than any feature on this list.

メリットとデメリット

メリット

  • Getting a working chatbot live takes minutes, and the embed is a single script tag for most sites.
  • Training sources sync automatically, so updated help articles flow into the agent without manual retraining.
  • Actions reach into Stripe, Shopify, Salesforce, and custom APIs, so the agent resolves issues instead of just describing them.
  • Escalation carries a transcript plus an AI summary, which saves human agents real time.
  • Multi-channel coverage from one agent spans web, email, WhatsApp, Slack, Instagram, and voice.

デメリット

  • Message credits, not conversations, are the billing unit, and one user question can consume several credits when an action fires, so costs are harder to predict than a flat per-seat price.
  • Training content is capped (1 MB on Free, 40 MB on Pro), which will bite anyone trying to load a large knowledge base.
  • Annual support tickets and auto-resync of websites and Notion pages are paid-plan features, so Free users have to update sources by hand.
  • Removing the "Powered by Chatbase" branding costs $99/month on top of your plan, which is a steep price for one small change.
  • Voice and phone support add complexity and cost, so teams should treat it as a step up rather than a default channel.

よくある質問

Chatbase is used to build AI customer support agents that answer questions from your own content and resolve issues across chat, email, WhatsApp, Slack, and phone. Teams use it to deflect repetitive tickets, collect leads, and escalate complex cases to humans with context attached.

関連コンテンツ

Chatbaseに関連するツール、スキル、記事を探す。

Chatbaseの代替ツール

Emailwritr

Emailwritr

Emailwritr · ライティング · マーケティング

Emailwritrは、すでに持っているリストからより多くの収益を引き出したいアフィリエイター、コース制作者、オンラインビジネス向けに作られたAIメールコピーライティングツールだ。空白のページや汎用的なチャットボットのプロンプトの代わりに、このAIメール作成ツールは、著名なダイレクトレスポンスのコピーライターを手本にしたフレームワークから、ウェルカムシーケンス、ローンチキャンペーン、カート放棄メール、プロモシリーズを生成する。自分のオートレスポンダーをつなぎ、オファーについて2、3の質問に答えると、あとは文案を書いてくれる。

無料 / $0 - $19.99/mo詳細を見る
Prosp

Prosp

Prosp · ライティング · 音声・言語 · マーケティング

Prospは、代理店や営業チーム向けに作られたAI LinkedInアプローチツールだ。見込み客ごとに自分の声でメッセージとボイスメッセージを書き、相手が本当に返信するように仕向ける。アカウントをつなぎ、リードを見つけ、AIにパーソナライズされたメッセージを大量に下書きさせて送信させる。すべて1つの受信トレイから行える。大量のアウトバウンドを回す人向けにできている。それがこの製品の売り込みのすべてだ。それでも各接点は人間らしく感じられなければならない。

有料 / $30.99 - $79.99 per account/mo詳細を見る
On Model

On Model

On Model · 画像 · マーケティング

On Modelは、服を仮想モデルに着せるAIファッションモデルツールであり、オンラインストアが撮影を予約せずにモデル着用の商品写真を手に入れられるようにする。Shopifyアプリとして動き、既存画像のモデルを差し替え、背景を変え、切り取られた写真や頭のない写真に顔を生成できる。狙いは単純だ。商品写真を顧客に合わせ、同じサプライヤー画像を使う競合と差をつける。

無料 / $0 - $20/mo詳細を見る