Dialogflow CX Review, Pricing & Alternatives (2026)

Dialogflow began as API.AI, a startup Google acquired in 2016, and became one of the most widely used tools for building chatbots and phone bots. Google offers two editions: Dialogflow ES, the original intent-based product, and Dialogflow CX, launched in 2020 for large, complex agents with a visual state-machine builder. In 2024 and 2025 Google folded CX into its Customer Engagement Suite and renamed the console Conversational Agents, with pricing unchanged for agents built there. It is used by banks, airlines, telecoms, retailers and governments to automate customer support across web chat, messaging apps and contact center phone lines.

Flows are the deterministic side. Designers map conversations visually as flows and pages, with intents trained on example phrases, entity extraction, form filling for collecting details such as account numbers, conditional routes, and webhooks that call back-end systems. This gives precise control for regulated tasks such as payments or identity checks. Agents support more than 100 languages and variants, versions and environments for testing and release, and built-in analytics. Voice agents connect to telephony through Google's Contact Center AI Platform and partners such as Genesys, Avaya and Twilio, with speech recognition, text-to-speech, barge-in and DTMF keypad input.

Playbooks are the generative side. Instead of hand-writing every path, builders describe goals and steps in natural language, add examples, and give the playbook tools such as APIs or data stores. Gemini models then plan responses and call tools. Data stores index websites, documents and structured data so the agent can answer questions with retrieval-augmented generation, and generators and generative fallback handle open-ended prompts or unmatched input. Hybrid agents call a flow from a playbook or a playbook from a flow, which lets teams keep strict logic where needed. Google also offers CX Agent Studio for Gemini agents with built-in evaluations, priced per session.

Billing is pay as you go on a Google Cloud account, with no subscription tiers. Flows cost $0.007 per chat request and $0.001 per second of voice audio, about $0.06 a minute. Playbooks cost $0.012 per chat request and $0.002 per voice second, about $0.12 a minute. Any turn that uses a playbook, data store, generator or generative fallback is billed at the Playbooks rate, and if a conversation starts in a playbook every turn is billed that way. Data store indexing includes 10 GiB free a month, then $5 per GiB. New users get trial credits of $600 for Flows and $1,000 for Playbooks, valid for 12 months.

Dialogflow CX's strengths are its mix of deterministic control and generative answers, mature voice and telephony support, broad language coverage and the scale of Google Cloud. Its weaknesses are a steep learning curve, per-request costs that are hard to forecast, and the need for developers to manage webhooks, environments and cloud setup. Amazon Lex is the AWS equivalent, and Microsoft Copilot Studio suits Microsoft 365 organizations. Kore.ai offers a broad enterprise platform, while Voiceflow and Botpress give designers faster visual building. Choose Dialogflow CX if you run a large contact center on Google Cloud and need precise flows alongside Gemini answers.