B

Botpress

Quebec City-built MIT-licensed open-source AI-agent platform — founded 2017 by Sylvain Perron (CEO) and Justin Watson, 14,686+ GitHub stars on botpress/botpress, ~$40M total funding including a $25M Series B (≈$34M CAD) on 23 June 2025 at $120M post-money valuation led by FRAMEWORK with Inovia Capital, Deloitte Ventures, HubSpot Ventures and Decibel; pivoted in 2023 from pure open-source toolkit to all-in-one LLM-agent platform with managed Cloud

AI Agents / Chatbot Platform / Customer Support 4.5 / 5 (1 Rating) Free (5K msg) / Plus $89 / Team $495 / Enterprise ~$2K/mo (plus AI tokens) Updated May 2026 🌍 Self-hostable open-source (RBI-friendly) + Cloud SaaS
✅ Recommended for Indian engineering-led customer-support and AI-agent teams

Quick Verdict

Botpress is one of the original and most credible open-source chatbot platforms — and as of 2024-2026, one of the most clearly positioned AI-agent infrastructure products for engineering-led teams. The company was founded in 2017 in Quebec City, Canada by Sylvain Perron (CEO) and Justin Watson (Jean-Bernard Perron is also part of the founding team). Botpress was open-sourced on GitHub in January 2017 under the MIT licence, and the main botpress/botpress repository now has 14,686+ GitHub stars. The company is headquartered in Quebec City with engineering teams in Montreal and a US presence in Delaware. Through five funding rounds Botpress has raised approximately $40 million, including a $15M Series A in 2021 at $50M post-money, and a headline $25M USD (≈$34M CAD) Series B on 23 June 2025 at $120M post-money valuation — led by FRAMEWORK with Inovia Capital, Deloitte Ventures, HubSpot Ventures, Decibel and existing investors. The single most important strategic event in the company's history was the 2023 pivot: Botpress moved from a pure open-source chatbot toolkit to an all-in-one cloud platform for building and deploying AI agents powered by LLMs (OpenAI GPT, Anthropic Claude, Google Gemini), with the original open-source server reframed as the self-host on-ramp into the Cloud product. For Indian customer-support, SaaS, fintech and IT-services teams in 2026 the right framing is: Botpress is the right call for engineering-led teams that want to build LLM-powered AI agents with code-level control, real RBAC, MIT-licensed self-host options for RBI / data-residency requirements, and a Cloud surface for teams that don't want to self-host. It is the wrong call for teams that need first-class Hindi / regional-Indian-language depth out of the box (use Yellow.ai or Haptik), or for non-engineering teams that want a fully no-code Intercom Fin-style assistant.

Engineering-led AI-agent building
4.7
Open-source self-hostability (MIT)
4.8
LLM provider flexibility
4.6
Indian-language / Hindi depth vs Yellow.ai
3.0
India support / regional presence
2.5

What is Botpress?

Botpress is a developer-focused platform for designing, building and deploying conversational AI agents. The platform's surface today is broader than the "chatbot" framing the original product launched with — Botpress is now positioned as infrastructure for autonomous LLM-driven agents that can handle customer support, internal IT helpdesk, sales qualification, lead routing, FAQ answering, RAG over documents, knowledge-base Q&A, and structured workflow execution (form fills, ticket creation, CRM updates) — with the LLM as the brain and Botpress as the surrounding execution + control layer.

The company was founded in 2017 in Quebec City, Canada by Sylvain Perron (a Quebec-based full-stack engineer who serves as CEO) and Justin Watson (co-founder). The project was open-sourced on GitHub in January 2017 under a permissive MIT licence, and rapidly grew a developer community — hitting 1,000 GitHub stars in 2018, 9,500+ stars by 2022, and 14,686+ stars as of 2026. The headquarters remain in Quebec City, with engineering teams in Montreal and a US legal entity in Delaware (the standard Canadian-startup raising-USD setup). The team is now several dozen strong across engineering, product, customer success and go-to-market.

Botpress's funding history is uneven and reflects the strategic pivot it went through. After several years of open-source community traction, the company raised a $15 million Series A in 2021 at a $50 million post-money valuation. The post-Series A period was spent building the cloud platform and pivoting positioning from "open-source chatbot toolkit" to "all-in-one platform for building AI agents powered by LLMs". The Botpress Cloud product launched in 2023, coinciding with the post-ChatGPT explosion in LLM-agent demand. Two years later, on 23 June 2025, the company closed a $25 million USD (≈$34 million CAD) Series B at a $120 million post-money valuation — more than 2x the Series A valuation, with FRAMEWORK leading and Inovia Capital, Deloitte Ventures, HubSpot Ventures and Decibel participating alongside existing investors. Total funding to date is approximately $40 million, putting Botpress in the same funding band as competitors like Voiceflow but well below Indian-domestic Yellow.ai (≈$100M+ raised) and Haptik (Reliance Jio-owned).

The 2023 pivot deserves a beat of analysis. Pre-2023, Botpress was the most popular self-hostable chatbot framework on GitHub but a small commercial business. The launch of the Cloud platform in 2023, on the same day as the wider GPT-4 enterprise wave, repositioned Botpress as a vendor-led AI-agent platform with an open-source self-host on-ramp — which is exactly the structure that has worked for n8n, GitLab, Sentry and other commercial-OSS players. By 2025, the company reported "thousands of customers running live agents in production" across customer support, fintech, IT services and consumer technology, and the Series B announcement framed the company as a core layer in the agentic AI stack.

What Botpress does (the product surface)

🤖 LLM-agnostic agent builder

Build agents that route between OpenAI GPT-4 / GPT-5, Anthropic Claude, Google Gemini, Mistral, and open-weight models. The LLM is pluggable — you can pick per-agent, per-skill, or fall back across providers. Important for Indian teams hedging against single-vendor LLM lock-in.

🧠 RAG / Knowledge Base

First-class retrieval-augmented generation over PDFs, websites, Notion pages, Google Drive, Confluence, Zendesk. Vector storage included in the platform (50 MB free, 1 GB on Plus, more on Team/Enterprise). Indian teams indexing product docs, policy manuals, and FAQs find this surface mature.

📞 Multi-channel deployment

Deploy the same agent across web widget, WhatsApp Business API, Messenger, Instagram, Telegram, Slack, Microsoft Teams, Twilio voice, custom HTTP API, and embedded in mobile apps. WhatsApp Business API is the channel that matters most for Indian buyers — covered well.

🔄 Workflows & tool calling

Agents can call APIs, query databases, look up CRM records, file Zendesk / Freshdesk tickets, and trigger Zapier / n8n / webhook workflows. Tool calling is implemented natively rather than as an afterthought — closer to the OpenAI Assistants API model than to legacy chatbot platforms.

👥 Human handoff & live agent inbox

Built-in human-in-the-loop: agents can escalate to live human support, route to specific queues, and human agents can pick up conversations mid-flow. Critical for Indian fintech / banking customer support where regulatory rules require human escalation for certain conversation types.

📊 Analytics, RBAC, SSO, audit logs

Enterprise tier adds SSO (SAML / OIDC), SCIM provisioning, audit logs, data-residency commitments, SOC 2, GDPR compliance. Critical for Indian BFSI / NBFC procurement and for SaaS companies selling to regulated buyers.

Pricing & plans (2026)

Botpress restructured pricing in May 2026 to bundle AI quotas into base plans (previously AI spend was billed entirely separately). Live rates from botpress.com/pricing:

  • Free5,000 messages/events per month, 1 bot, 50 MB vector DB, 1 collaborator, "Powered by Botpress" branding. Genuinely usable for an Indian developer or small startup testing AI-agent ideas.
  • Pay-as-you-go$0.01 per message on top of the Free plan if you exceed 5,000/month. Unlimited bots and storage at this volume. Simple model for unpredictable traffic.
  • Plus$89/month + AI spend. Includes 1 GB Vector DB, human handoff for live support, AI conversation insights (sentiment analysis), and the ability to remove "Powered by Botpress" branding. Best for: Indian solo founders and pre-revenue SaaS startups.
  • Team$495/month ($445/month annual). Includes 50,000 messages, 3 bots, live agent handoff, priority support, real-time collaboration. The most-bought tier for Indian Series A+ SaaS / fintech / IT-services teams.
  • Enterprise — custom pricing, typically starting around $2,000/month on 3-year contracts. Adds SSO, SCIM, audit logs, data residency commitments, SLAs, dedicated CSM, and elevated compliance posture (SOC 2, GDPR). Best for: Indian BFSI / fintech / large-enterprise buyers.
  • Self-hosted (open-source)MIT-licensed Botpress server, free forever. You run it on your own infrastructure (Docker / Kubernetes). No commercial support, no SSO, no audit logs — but full RBI / data-residency-friendly deployment surface for Indian banks and NBFCs that can't put conversation data in foreign cloud regions.

Important cost reality for Indian buyers: AI model tokens are billed on top of every paid plan (and on top of self-hosted). High-volume Indian customer-support deployments using GPT-4o or Claude Sonnet often incur several hundred to several thousand USD/month in LLM token spend beyond the Botpress subscription itself. For a typical Indian SaaS team on Team tier with 50K messages/month using GPT-4o for ~30% of conversations, the all-in monthly cost typically clusters at $700–$1,500/month (~₹60K–₹1.3L) with the Botpress subscription representing the smaller share.

When Botpress is the right call

  1. You're an Indian engineering-led customer-support, fintech or SaaS team building AI agents in production — Botpress is calibrated for teams with at least one engineer who can wire up tool calls, write custom skills, and integrate the agent with the rest of the stack. The LLM-agnostic architecture matters as Indian teams hedge between OpenAI, Anthropic, and Indian-hosted open-weight options.
  2. You need MIT-licensed self-host for RBI / data-residency compliance — Indian banks, NBFCs and healthcare-tech players with strict on-shore-data requirements can deploy the open-source server on AWS Mumbai or Indian on-premises infrastructure and never send conversation data outside India. Very few competitors offer this.
  3. You're an Indian SaaS company selling to enterprise buyers that demand SSO / SCIM / SOC 2 — the Enterprise tier covers the procurement-checklist requirements that come up in Indian BFSI / large-enterprise RFPs.
  4. You want to build agents that actually execute work, not just answer FAQs — tool calling, workflow orchestration, ticket-creation and CRM updates are first-class. Closer to building an autonomous agent than a scripted FAQ tree.

Botpress is the wrong call when: you need deep Hindi / regional-Indian-language depth with India-specific language modelling out of the box (use Yellow.ai or Haptik, both Indian-built with serious Hindi / Tamil / Marathi / Bengali capability); you need a fully no-code Intercom Fin-style assistant with zero engineering effort (use Intercom Fin / Drift / Zendesk Answer Bot); you're a Reliance Jio ecosystem company (Haptik is the obvious choice given Jio ownership); or you don't have any engineering capacity (Botpress's Cloud is friendlier than the OSS but still benefits meaningfully from a developer doing the integration work).

Pros & cons

✓ Pros

  • Genuine MIT-licensed open-source — self-hostable and forkable
  • 14,686+ GitHub stars; one of the most popular OSS chatbot platforms
  • LLM-agnostic architecture — pluggable across OpenAI / Anthropic / Gemini / open-weight
  • Tool calling, RAG, multi-channel deployment all first-class
  • ~$40M funded with $25M Series B (June 2025, FRAMEWORK led) — financially stable
  • Founder team is technical (Sylvain Perron is a full-stack engineer first)
  • RBI / data-residency-friendly via OSS self-host
  • Generous Free tier — 5K messages and 50MB vector DB
  • WhatsApp Business API support — critical for Indian buyers
  • Enterprise tier covers SSO / SCIM / SOC 2 / GDPR / audit logs

✗ Cons

  • Hindi / regional-Indian-language depth lags Yellow.ai and Haptik
  • AI model tokens billed separately — total cost can surprise teams
  • Best with at least one developer on the team — not zero-engineering
  • Cloud pricing in USD with 18% IGST; no INR billing option
  • No India-region cloud or IST-aligned support
  • Smaller Indian community and partner ecosystem than Yellow.ai
  • Self-hosted OSS lacks enterprise features (SSO, SCIM) — Cloud-only for those
  • Strategic pivot from OSS-only to platform created some community friction (visible on GitHub issues circa 2023-2024)

Related insights & playbooks