Voice agent platforms compared
A voice agent platform runs the call and joins the parts together. Most charge a platform fee per minute and bill speech, the language model and the phone line on top, so the headline rate is rarely what a call costs.
Too few independent figures to compare these side by side yet. List prices are on the second tab.
- Single figures
- 0 of 10Each from a different test, not comparable
- Vendor claims
- 6 of 10Publish their own figure
- Lowest price
- $0.01Daily Pipecat Cloud
- Highest price
- $0.14Bland AI
10 compared, read 29 October 2026 or later. Click a bar to open its row, a column heading to sort, ↗ for the source. Tick Compare on up to four rows.
Showing 10 of 10
No independent performance figures yet
Too few to compareNo independent source has measured enough of these 10 the same way to chart them side by side. Below: the single figures there are, who measures them elsewhere, and 6 vendor claims, folded and unranked. Price is on the next tab.
Measured elsewhere
No independent comparison found.
What RingTru will measure
The wait from the moment a caller stops talking to the first word of the reply, timed the same way on every platform over a UK phone line.
What each platform says about its speed (6, unranked)
Each vendor’s own words, in name order. They time different things, so they are not charted or ranked.
- Bland AI Under 400 ms response latency docs.bland.ai
- ElevenLabs ElevenAgents 620 ms time to first audio for the Realtime v1 Mini speech-to-speech model (570 ms for Realtime v1 Max), no figure for the standard speech-to-text, model, text-to-speech pipeline elevenlabs.io
- LiveKit Cloud 100 ms global latency (network transport figure, not voice-agent reply time) livekit.com
- Retell AI About 600 ms retellai.com
- Synthflow Sub-500 ms (voice agents), telephony network target under 100 ms round trip synthflow.ai
- Vapi Typically around 800 ms end to end docs.vapi.ai
- Platform fee, lowestDaily Pipecat Cloud $0.01
Bar length is the figure itself, so the shortest bar is best in every column.
| US$ a minute | Checks✓ yes, ✕ no, ~ partly | |
|---|---|---|
| Vapi Vapi | $0.05 | UK numbers Own models |
| Retell AI Retell AI | $0.055 | UK numbers Own models |
| ElevenLabs ElevenAgents (ElevenLabs Agents, formerly Conversational AI) | $0.08 | Own models |
| Bland AI Bland | $0.14 | Own models |
| Synthflow Synthflow | Not published | UK numbers |
| LiveKit LiveKit Cloud (Agents) | $0.01Lowest | UK numbers Own models |
| Daily Pipecat Cloud | $0.01Lowest | Own models |
| Deepgram Voice Agent API | $0.075 | UK numbers Own models |
| Twilio ConversationRelay | $0.07 | UK numbers Own models |
| Ultravox (Fixie.ai) Ultravox Realtime | $0.05 | Own models |
Vendor claims, data regions and compliance are in each row’s Details, in the vendor’s own words. Claims are never charted or ranked. RingTru has measured none of these yet. Add up a whole call in the calculator.
How to read these numbers
- The platform fee is only part of the bill
- Most platforms list a per-minute fee for running the call, then pass through speech-to-text, the language model, text-to-speech and the phone line, each billed separately. Add the layers up with the calculator on the arena page before comparing.
- Latency claims measure different things
- One platform quotes the time from the caller stopping to the agent starting, another quotes its own processing time alone. Unless two figures name the same start and end points, they cannot be compared.
- UK numbers
- Of the ten platforms here, only Twilio sells UK numbers itself. Five say a UK number has to be bought from a carrier such as Twilio or Telnyx and then imported or connected over SIP, and four publish nothing we could find about UK numbers. Either way the carrier brings its own UK rules and prices, listed on the phone carriers page.
Go deeper
Your agent, tested on real calls.
- test calls
- 506
- passed
- 96.8%
- median reply
- 0.94 s
Holds the new day when a caller changes their mind.