Voiceflow in 2026: The Build-Your-Own Agent Option
Confirmed with Voiceflow, 20 August 2026: categorisation, product description and pricing framing. Customer logos and usage counts are the figures Voiceflow publishes. Voiceflow does not publish a company headcount or named plan tiers, so neither appears here. Last updated 21 August 2026.
Voiceflow is a different kind of product from the finished support bots elsewhere on this site. It is a build-your-own agent platform: you design and control the agent logic, the LLM, and the responses, rather than buying a finished resolution engine that is priced per resolution. If you want a bot to switch on next week, Voiceflow is not it. If you want to own how the agent works, it is exactly the category you are looking at.
What Voiceflow Is
Voiceflow is a platform for building, orchestrating, and controlling your own AI agent across chat and voice. Rather than shipping a done resolution engine, it gives a support team the tools to assemble the agent themselves, choosing the logic, the LLM, and the responses. It works across web chat, voice, and its API, and connects to multiple LLM providers.
It is a Toronto platform publishing 4,000+ customers and 200,000+ users, with enterprise logos including JPMorgan Chase, Optum, Allstate, Cisco, Vodafone, Turo, StubHub and Rocket. How much implementation falls to you depends on the route: support is self-serve or fully managed.
Pricing (Contact Sales)
Free trial with no credit card required, then usage-based billing. No prices are published on voiceflow.com/pricing (Request pricing / Book a demo); business pricing is quoted.
Deflection
Voiceflow publishes no platform-wide deflection rate, and on a build-your-own platform a single number would mislead: the result follows your own design work. The one figure attributable to a named case study is Trilogy, which runs support for 90+ products and reports 70% of tickets resolved by AI, up from 35% in week one and 59% at twelve weeks. Read it as what a well-designed build can reach, not as a rate to expect by default.
Source: Voiceflow customer story: Trilogy (vendor-published, last updated May 2026).
Implementation Time
Days to weeks if you build it yourself. Implementation support is self-serve or fully managed, so the build can also be run for you.
- +CX teams that want to build and control their own agent across chat and voice
- +Teams that want to own the logic, LLM, and responses rather than buy a finished engine
- +Buyers comfortable with a contact-sales, usage-based platform
- -Teams that want a finished bot to switch on
- -Buyers who need a published per-resolution price
- -Buyers who want a fixed per-resolution price rather than usage-based billing
How Voiceflow Differs From the Finished Bots Here
Most vendors compared on this site are finished bots. You point them at your knowledge base, connect a helpdesk, and they resolve tickets, usually priced per resolution or per conversation (for example, HubSpot Breeze at $0.50 per resolution or Intercom Fin at $0.99 per resolution). The pricing is public and the agent behaviour is largely the vendor's.
Voiceflow is the opposite trade. You build the agent, so you carry the design work, but you also own the logic, the model choice, and the responses. That is why there is no published per-resolution price: pricing is contact-sales and usage-based, and the outcome depends on what you build rather than on a fixed resolution engine.
Compare Finished Bots
Questions About Voiceflow
How much does Voiceflow cost?
Voiceflow offers a free tier plus Pro, Business, and Enterprise plans, but no prices are published on voiceflow.com/pricing. The paid tiers are shown as Request pricing or Book a demo, and Enterprise is quote-only. Pricing is usage-based and contact-sales.
Is Voiceflow a finished customer service bot?
No. Voiceflow is a build-your-own agent platform. You design and control the agent logic, the LLM, and the responses yourself, rather than buying a finished resolution engine that is switched on and priced per resolution.
Who is Voiceflow for?
CX teams that want to build, orchestrate, and control their own AI agent across chat and voice, rather than buy a finished bot. It is not for teams that want a finished bot to switch on, or a published per-resolution price.