LiveKit
- Strength
- Ranks third on repeatable reliability with 99.19% infrastructure-clean calls.
- What can be improved
- Consent was collected, but consent_id was omitted from the handoff tool.
- nova-3
- openai/gpt-4.1
- sonic-3
Retell is ahead on repeatable reliability, task completion, infrastructure reliability, tool call accuracy and response time, and LiveKit on voice tone and clarity and interruption handling.
Last updated Methodology by Luis Ojeda
Pick LiveKit for how the agent sounds (0.04 higher) and handling interruptions (0.19 higher). Pick Retell for reliability across repeated runs (19.51 points higher), completing the caller's task (2.44 points higher), calls that connect and stay up (0.81 points higher), accurate tool calls (0.28 higher) and fast replies (1.64s faster). For building on LiveKit, see our guide to LiveKit agents.
Bars share one scale across all 10 platforms. The place next to each figure is its rank in the field.
Share of the 82 scenarios that passed on all three runs.
Retell, 19.51 points higher
Calls where the expected outcome was fully reached.
Retell, 2.44 points higher
Calls with no connection, audio or platform failure.
Retell, 0.81 points higher
Mean score for calling the right tool with the right arguments.
Retell, 0.28 higher
Mean score for how clear and natural the agent sounds.
LiveKit, 0.04 higher
Mean score for yielding and recovering when the caller cuts in.
LiveKit, 0.19 higher
Mean time for the agent to start replying after the caller stops.
Retell, 1.64s faster
246 calls per platform: 82 scenarios, 3 runs each.
Retell is more reliable, by 19.51 points. LiveKit passed 70.73% of the 82 scenarios on all 3 runs and Retell passed 90.24%. A scenario only counts when every run of it passed.
Retell is more reliable, by 19.51 points. LiveKit passed 70.73% of the 82 scenarios on all 3 runs and Retell passed 90.24%. A scenario only counts when every run of it passed.
Retell starts replying sooner, 1.64s faster. Mean response time is 2.59s for LiveKit and 0.95s for Retell.
LiveKit reached the expected outcome on 95.12% of calls and Retell on 97.56%. Retell is ahead on task completion.
Both ran the same Appointment and Medicare agents through the same 82 caller scenarios, 3 times each, with the same evaluators and mock tools. Only the platform and its speech and model components changed.
Cekura runs these simulated calls on your agent and scores every one.