Botpress vs Fin by Intercom

Botpress is a platform for building AI agents and chatbots powered by large language models. The Canadian company began with an open-source chatbot framework and later rebuilt the product as a cloud platform centered on a visual studio, where builders combine autonomous LLM nodes with structured workflows and can drop into JavaScript code whenever a step needs custom logic.

Agents draw on knowledge bases built from websites, documents and tables, call external systems through more than 100 integrations, and deploy to web chat, WhatsApp, Messenger, Telegram, Slack, Microsoft Teams and other channels. Human handoff, analytics, versioning and role-based collaboration are layered across plans. Pricing starts with a free plan of 100 conversations, then Plus and Team plans priced by included monthly conversations, with extra conversation packs and an AI usage allowance; usage beyond that allowance is billed, so the total cost depends on traffic and the models you choose.

Botpress fits developers, agencies and product teams building several agents that must follow defined processes and connect to real systems. Non-technical users who simply want a website bot trained on their content will be faster with Chatbase, Voiceflow offers a more design-oriented collaboration experience for conversation teams, and support organizations already on a help desk may prefer Intercom Fin or Zendesk AI agents. The product evolves quickly, so older tutorials and community answers can go out of date.

Fin is the AI customer service agent from Intercom, the Dublin- and San Francisco-based company founded in 2011 that popularized in-app messaging. Launched in 2023, Fin answers customer questions in chat, email and other channels by drawing on help center articles, internal documentation, past conversations and connected data sources, and it can take actions such as looking up orders or updating accounts through API connectors.

Fin works inside Intercom's own help desk and as a standalone agent for other help desks including Zendesk, Salesforce Service Cloud and HubSpot, so teams can add it without migrating. Procedures and guidance let support managers define how Fin should behave, which policies to follow and when to hand off to a human. It works in many languages, including Turkish, and reports show resolution rate, topics and gaps in content. Pricing is usage-based, charged per successful outcome rather than per seat, which aligns cost with results but makes spend rise with volume.

Fin suits support teams with meaningful volume and good documentation, where a large share of questions repeat. Teams with thin knowledge bases or very low volume see less benefit. Zendesk AI agents, Salesforce Agentforce and Decagon or Sierra are competing approaches, and Chatwoot's Captain or Crisp's AI features are cheaper options for smaller teams.

Intercom now presents itself primarily as an AI-first company, and successive Fin releases since 2024 have added voice support, deeper actions and better analytics.