ElevenAgents by ElevenLabs Review

ElevenAgents (formerly the ElevenLabs Conversational AI / Agents Platform) runs real-time voice and text agents. Each agent chains a speech-to-text model, a large language model of your choice (a supported third-party model or your own custom LLM), an ElevenLabs text-to-speech voice and a turn-taking model that handles pauses and interruptions. You configure the system prompt, voice, language, knowledge base (RAG), tools that call your APIs, and multi-step workflows in a dashboard or visual workflow builder, or manage agents as code with a CLI and API.

Agents deploy as an embeddable web widget, through SDKs for React, Swift, Kotlin, React Native and Unity, over WebSocket, or on the phone through native Twilio, SIP trunking and batch outbound calling. ElevenLabs also lists WhatsApp, SMS, email and chat channels, native connectors such as Salesforce, HubSpot, Zendesk, Stripe and Shopify, plus automated tests, simulations, A/B experiments, transcript search and an analytics dashboard.

ElevenAgents uses the ElevenLabs subscription tiers, shown on a separate agents pricing page. Every paid plan's included minutes work out to $0.08 a minute.

Usage rates published on the same page, identical on every tier: additional call minutes $0.08 a minute, burst calls $0.16 a minute, text messages $0.003 each. LLM usage is billed on top, based on the model you pick, with the fee shown in the platform and deducted from your ElevenLabs credits. ElevenLabs charges no telephony fee, but if you connect Twilio or a SIP trunk, that carrier bills you directly. Burst pricing, when enabled, lets an agent take up to 3 times your concurrency limit (capped at 300 for non-enterprise accounts) at double the minute rate, with lower processing priority. There is no limit on the number of agents. Startups can apply for a grant of 12 months free. Prices exclude taxes.

Prices checked on the vendor's pricing page on November 1, 2026. See elevenlabs.io/pricing/agents.

What ElevenLabs' own pages and docs list:

ElevenAgents suits teams for whom the voice is part of the product: customer support lines, appointment booking, sales qualification and branded assistants where callers hear the agent. It fits developers who want a hosted pipeline with ElevenLabs voices, a choice of LLM, and published per-minute pricing they can start on for $6.

If you want more control over each component and carrier, Vapi is the developer-first alternative. Retell AI and Synthflow target phone automation with built-in telephony workflows, and PolyAI sells managed enterprise contact-centre agents. If you only need speech for your own agent stack, see Stork's main ElevenLabs page for the TTS and STT API.

Reviewers mostly agree the voice is the strength and the bill is the weakness. Layer3Labs, an agency that builds voice-agent systems, calls it one of the best-sounding voice layers for live calls but warns "the headline rate covers only part of the total cost" because LLM and telephony are billed separately, and that ElevenLabs publishes no turn-taking accuracy figures. Retell AI, a direct competitor, likewise rates its voices top of the category while citing G2 reviewers who find production monitoring thin and noting HIPAA BAAs and data residency are enterprise-only. CloudTalk, which sells its own AI voice agents, lists credit consumption on long calls and helpdesk integrations that need engineering work among the main complaints.