Skip to content

Roark Live Bench · Suite v1.3 · Sweep of 20 November 2026

Voice AI models, ranked on live phone calls.

Every speech-to-text, language model and text-to-speech model, run in every combination on the same real call to the same agent. A model is ranked on every stack it runs in; the stacks are the evidence.

1234 stacks published24,562 scored callsPublished at 20 scored calls

Our picks

The read

  • 505 of 916 cascades complete the task in at least 90% of calls, and the top 10 sit within 1 point of each other. What separates the top stacks is speed and cost.
  • Median response time runs from 1.2 s to 6.2 s on the same agent and the same caller, a spread of 5.0 s that comes entirely from the models behind it.
  • No stack wins everything. The fastest, Speechmatics Linden-1 (own VAD) + Anthropic Claude Haiku 4.5 + AssemblyAI TTS-1 at 1.2 s, ranks 855 of 916 on task completion; the most complete is Speechmatics Linden-1 (own VAD) + OpenAI GPT-6.1 Sol + OpenAI TTS-1 HD at 99%.
  • The language model matters most: swapping it moves task completion by 9 points, against 7 for the transcriber and 5 for the speech engine.

Cascade vs Speech-to-Speech

The first choice to make. A cascade chains three models: one transcribes, one decides what to say, one speaks. A speech-to-speech model does all three at once. Both run the same agent on the same calls here.

  • Cascades lead on all 6 measures compared here.
  • Speech-to-speech costs 3.7 times as much per completed task: $0.24 for speech-to-speech against $0.07 for cascades.
CascadeSpeech-to-speech
Task completionTool useTime to first wordResponse timeBarge-in handlingCost per task255075100
MeasureCascadeSpeech-to-speech
Task completion
92% (leads)85%
Tool use
87% (leads)78%
Time to first word
0.6 s (leads)1.1 s
Response time
2.1 s (leads)2.1 s
Barge-in handling
100% (leads)96%
Cost per task
$0.07 (leads)$0.24

The three strongest of each, averaged. Cascades: Deepgram Flux + OpenAI GPT-6.1 Terra + AssemblyAI TTS-1; Deepgram Flux + Meta Muse Ember 1.0 + AssemblyAI TTS-1; Speechmatics Linden-1 (own VAD) + OpenAI GPT-6.1 Terra + AssemblyAI TTS-1. Speech-to-speech: grok-voice-think-fast-2.0; gpt-live-1 (backend gpt-6-luna); grok-voice-3.0. Each measure is scored 0 to 100 on its own scale, further out is better; the table gives the same figures in their own units.

Which to use

  • Choose speech-to-speech when the pause matters more than the outcome: short transactional calls, callers who interrupt, queues where every half second costs you callers. Expect to pay more, and choose the model carefully (grok-voice-3.0 leads here).
  • Choose a cascade when the call has to complete, when cost matters, or when you need to see which stage to fix. It is 12 points ahead on task completion here.
  • Costs are what each stack’s providers charged for these calls, at the rates in effect when they ran. Telephony and hosting are not included.

Speech-to-Text

The transcriber turns the caller's speech into text. It decides what the agent thinks it heard, and nothing after it can start until it finishes.

  • Soniox v6 and AssemblyAI-STT-3.7 Realtime-Agent are level: neither leads by more than the margin of error. Of the two, Soniox v6 costs least, at $0.008 a call.
  • Deepgram Flux delivers its transcript in 0.38 s, OpenAI Whisper-1 in 1.93 s: 1.54 s added to every turn of the call.
  • Soniox v6 is the most accurate: it changes the meaning of 0.2% of the caller's words, against 2% for Speechmatics Linden-1 (own VAD).
  • In 587 of 1,218 cascades the transcriber is the slowest step of the turn, whichever model runs it.

Roark score, best first

100Soniox v697AssemblyAI-STT-3.7 Realtime-Agent92Universal-3.6 Pro84Nova-382Universal-3.5 Pro73Enhanced (realtime)70Flux69Nova-269Linden-167Scribe v3 Live64Standard (realtime)57Linden-1 (own VAD)55Whisper-151GPT Transcribe
Every transcriber on every figure, sorted by roark score. Column headers sort the table.
#
1Soniox v6Soniox100 / 100. Semantic word error rate 0.2%, 71.4 of 71 points; Time to final transcript 0.85 s, 28.6 of 29 points(among the best)0.2%(among the best)0.85 s(among the best)$0.008(among the best)
2AssemblyAI-STT-3.7 Realtime-AgentAssemblyAI97 / 100. Semantic word error rate 0.6%, 68.1 of 71 points; Time to final transcript 0.86 s, 28.6 of 29 points(among the best)0.6%(among the best)0.86 s$0.030
3Universal-3.6 ProAssemblyAI92 / 100. Semantic word error rate 0.8%, 63.7 of 71 points; Time to final transcript 0.90 s, 28.6 of 29 points(among the best)0.8%(among the best)0.90 s$0.029
4Nova-3Deepgram84 / 100. Semantic word error rate 1.1%, 55.4 of 71 points; Time to final transcript 0.88 s, 28.6 of 29 points1.1%0.88 s$0.031
5Universal-3.5 ProAssemblyAI82 / 100. Semantic word error rate 1.1%, 53.9 of 71 points; Time to final transcript 0.90 s, 28.6 of 29 points1.1%0.90 s$0.029
6Enhanced (realtime)Speechmatics73 / 100. Semantic word error rate 1.4%, 44.4 of 71 points; Time to final transcript 0.98 s, 28.6 of 29 points1.4%0.98 s$0.054
7FluxDeepgram70 / 100. Semantic word error rate 1.6%, 41.4 of 71 points; Time to final transcript 0.38 s, 28.6 of 29 points1.6%0.38 s(among the best)$0.031
8Nova-2Deepgram69 / 100. Semantic word error rate 1.6%, 40.3 of 71 points; Time to final transcript 0.89 s, 28.6 of 29 points1.6%0.89 s$0.023
9Linden-1Speechmatics69 / 100. Semantic word error rate 1.6%, 40.2 of 71 points; Time to final transcript 0.98 s, 28.6 of 29 points1.6%0.98 s$0.020
10Scribe v3 LiveElevenLabs67 / 100. Semantic word error rate 1.7%, 38.4 of 71 points; Time to final transcript 0.88 s, 28.6 of 29 points1.7%0.88 s$0.015(among the best)
11Standard (realtime)Speechmatics64 / 100. Semantic word error rate 1.8%, 35.6 of 71 points; Time to final transcript 0.96 s, 28.6 of 29 points1.8%0.96 s$0.030
12Linden-1 (own VAD)Speechmatics57 / 100. Semantic word error rate 2%, 28.2 of 71 points; Time to final transcript 0.46 s, 28.6 of 29 points2%0.46 s(among the best)$0.020
13Whisper-1OpenAI55 / 100. Semantic word error rate 1.5%, 43.6 of 71 points; Time to final transcript 1.93 s, 10.9 of 29 points1.5%1.93 s$0.025
14GPT TranscribeOpenAI51 / 100. Semantic word error rate 1.9%, 31.0 of 71 points; Time to final transcript 1.47 s, 19.5 of 29 points1.9%1.47 s$0.018(among the best)

Open a model to see where it ranks and the stacks it works best in.

Cost against qualitySemantic word error rate · Cost per call
$0$0.02$0.04$0.06$0.080%1%2%3%Cheap and accurateAccurate, pricierCheap, less accuratePricier and less accurateSoniox v6Scribe v3 LiveGPT TranscribeLinden-1 (own VAD)Linden-1Nova-2Whisper-1Universal-3.6 ProUniversal-3.5 ProAssemblyAI-STT-3.7 Realtime-AgentStandard (realtime)FluxNova-3Enhanced (realtime)Cost per call · ← betterSemantic word error rate · better →
Speed against qualitySemantic word error rate · Time to final transcript
0 s1 s2 s3 s0%1%2%3%Fast and accurateAccurate, slowerFast, less accurateSlower and less accurateFluxLinden-1 (own VAD)Soniox v6AssemblyAI-STT-3.7 Realtime-AgentScribe v3 LiveNova-3Nova-2Universal-3.5 ProUniversal-3.6 ProStandard (realtime)Linden-1Enhanced (realtime)GPT TranscribeWhisper-1Time to final transcript · ← betterSemantic word error rate · better →

Language model

The language model decides what the agent says and does. It has the most influence on whether the task gets done, and usually takes the largest share of each turn.

  • OpenAI GPT-6.2 Sol, OpenAI GPT-6.1 Terra and OpenAI GPT-6.1 Sol are level: none leads by more than the margin of error. Of them, OpenAI GPT-6.1 Terra costs least, at $0.013 a call. Avoid Google Gemini 4 Neon: last on task completion (81%), instruction following (3.8 / 5) and time to first token (2.67 s).
  • Anthropic Claude Haiku 4.5 starts answering in 0.36 s, Google Gemini 4 Neon in 2.67 s: 2.32 s added to every turn of the call.
  • OpenAI GPT-6.2 Sol uses its tools most reliably: it handles its tools without a mistake on 86% of calls, against 42% for OpenAI GPT-5.1.
  • In 378 of 1,218 cascades the language model is the slowest step of the turn, whichever model runs it.

Roark score, best first

66GPT-6.2 Sol65GPT-6.1 Terra63GPT-6.1 Sol57Claude Sonnet 5.557Muse Ember 1.056Claude Haiku 5.549Gemini 3.8 Flash47GPT-4.145Claude Haiku 4.543Gemini 4 Hydrogen43GPT-4o mini42GPT-4o40GPT-5.120Gemini 4 Neon

Not ranked here for missing data: Groq GPT-OSS 120B (no task completion or tool use or instruction following or time to first token).

Every language model on every figure, sorted by roark score. Column headers sort the table.
#
1GPT-6.2 SolOpenAI66 / 100. Task completion 91%, 20.3 of 36 points; Tool use 86%, 19.9 of 27 points; Instruction following 4.3 / 5, 11.8 of 18 points; Time to first token 0.87 s, 13.7 of 18 points(among the best)91%(among the best)4.3 / 5(among the best)86%(among the best)0.87 s67%$0.025
2GPT-6.1 TerraOpenAI65 / 100. Task completion 90%, 19.0 of 36 points; Tool use 86%, 19.8 of 27 points; Instruction following 4.3 / 5, 11.6 of 18 points; Time to first token 0.83 s, 14.2 of 18 points(among the best)90%4.3 / 5(among the best)86%(among the best)0.83 s67%$0.013
3GPT-6.1 SolOpenAI63 / 100. Task completion 92%, 22.0 of 36 points; Tool use 82%, 17.5 of 27 points; Instruction following 4.3 / 5, 11.8 of 18 points; Time to first token 1.00 s, 12.1 of 18 points(among the best)92%(among the best)4.3 / 5(among the best)82%(among the best)1.00 s72%$0.018
4Claude Sonnet 5.5Anthropic57 / 100. Task completion 91%, 19.9 of 36 points; Tool use 75%, 13.6 of 27 points; Instruction following 4.2 / 5, 11.0 of 18 points; Time to first token 0.93 s, 12.9 of 18 points91%(among the best)4.2 / 575%0.93 s–$0.114
5Muse Ember 1.0Meta57 / 100. Task completion 90%, 18.4 of 36 points; Tool use 67%, 9.2 of 27 points; Instruction following 4.3 / 5, 11.8 of 18 points; Time to first token 0.56 s, 17.5 of 18 points90%4.3 / 5(among the best)67%0.56 s85%(among the best)$0.005(among the best)
6Claude Haiku 5.5Anthropic56 / 100. Task completion 90%, 17.3 of 36 points; Tool use 69%, 10.2 of 27 points; Instruction following 4.3 / 5, 12.2 of 18 points; Time to first token 0.63 s, 16.5 of 18 points90%4.3 / 5(among the best)69%0.63 s–$0.007(among the best)
7Gemini 3.8 FlashGoogle49 / 100. Task completion 92%, 21.2 of 36 points; Tool use 69%, 10.4 of 27 points; Instruction following 4.3 / 5, 11.5 of 18 points; Time to first token 1.53 s, 5.7 of 18 points92%(among the best)4.3 / 5(among the best)69%1.53 s–$0.015
8GPT-4.1OpenAI47 / 100. Task completion 89%, 17.1 of 36 points; Tool use 52%, 1.2 of 27 points; Instruction following 4.2 / 5, 11.2 of 18 points; Time to first token 0.52 s, 18.0 of 18 points89%4.2 / 552%0.52 s82%(among the best)$0.016
9Claude Haiku 4.5Anthropic45 / 100. Task completion 88%, 14.3 of 36 points; Tool use 52%, 0.9 of 27 points; Instruction following 4.3 / 5, 11.4 of 18 points; Time to first token 0.36 s, 18.2 of 18 points88%4.3 / 5(among the best)52%0.36 s(among the best)–$0.052
10Gemini 4 HydrogenGoogle43 / 100. Task completion 91%, 20.9 of 36 points; Tool use 70%, 11.0 of 27 points; Instruction following 4.0 / 5, 9.4 of 18 points; Time to first token 1.88 s, 1.4 of 18 points91%(among the best)4.0 / 570%1.88 s–$0.049
11GPT-4o miniOpenAI43 / 100. Task completion 86%, 10.1 of 36 points; Tool use 56%, 3.3 of 27 points; Instruction following 4.2 / 5, 11.1 of 18 points; Time to first token 0.50 s, 18.2 of 18 points86%4.2 / 556%0.50 s(among the best)82%(among the best)$0.001(among the best)
12GPT-4oOpenAI42 / 100. Task completion 86%, 11.0 of 36 points; Tool use 53%, 1.8 of 27 points; Instruction following 4.2 / 5, 11.2 of 18 points; Time to first token 0.45 s, 18.2 of 18 points86%4.2 / 553%0.45 s(among the best)–$0.021
13GPT-5.1OpenAI40 / 100. Task completion 87%, 12.9 of 36 points; Tool use 42%, 0.0 of 27 points; Instruction following 4.1 / 5, 10.4 of 18 points; Time to first token 0.59 s, 17.1 of 18 points87%4.1 / 542%0.59 s74%$0.026
14Gemini 4 NeonGoogle20 / 100. Task completion 81%, 1.2 of 36 points; Tool use 70%, 10.7 of 27 points; Instruction following 3.8 / 5, 7.7 of 18 points; Time to first token 2.67 s, 0.0 of 18 points81%3.8 / 570%2.67 s–$0.078

Open a model to see where it ranks and the stacks it works best in.

Cost against qualityTask completion · Cost per call
$0$0.05$0.1$0.1575%80%85%90%95%Cheap and reliableReliable, pricierCheap, less reliablePricier and less reliableGPT-4o miniMuse Ember 1.0Claude Haiku 5.5GPT-6.1 TerraGemini 3.8 FlashGPT-4.1GPT-6.1 SolGPT-4oGPT-6.2 SolGPT-5.1Gemini 4 HydrogenClaude Haiku 4.5Gemini 4 NeonClaude Sonnet 5.5Cost per call · ← betterTask completion · better →
Speed against qualityTask completion · Time to first token
0 s1 s2 s3 s4 s75%80%85%90%95%Fast and reliableReliable, slowerFast, less reliableSlower and less reliableClaude Haiku 4.5GPT-4oGPT-4o miniGPT-4.1Muse Ember 1.0GPT-5.1Claude Haiku 5.5GPT-6.1 TerraGPT-6.2 SolClaude Sonnet 5.5GPT-6.1 SolGemini 3.8 FlashGemini 4 HydrogenGemini 4 NeonTime to first token · ← betterTask completion · better →

Text-to-Speech

The speech engine is the voice the caller hears, and the last step before they hear anything.

  • AssemblyAI TTS-1, Cartesia Sonic-3.6 and Cartesia Sonic-3.5 are level: none leads by more than the margin of error. Of them, AssemblyAI TTS-1 costs least, at $0.052 a call. Avoid OpenAI TTS-1 HD: last on readback accuracy (81%), word error rate (7.4%) and time to first audio byte (2.09 s).
  • AssemblyAI TTS-1 starts speaking in 0.04 s, OpenAI TTS-1 HD in 2.09 s: 2.04 s added to every turn of the call.
  • AssemblyAI TTS-1 is also the most accurate: it reads codes, amounts and dates back correctly 90% of the time, against 81% for OpenAI TTS-1 HD.
  • Rime Coda sounds the most human: 3.5 / 5, against 3.2 / 5 for ElevenLabs Turbo v2.5.
  • In 119 of 1,218 cascades the speech engine is the slowest step of the turn, most often with OpenAI TTS-1 (65 of them).

Roark score, best first

84TTS-183Flux TTS83Sonic-3.683Sonic-3.583v4 Turbo79Sonic-376Turbo v2.576Flash v2.574Coda58TTS-144TTS-1 HD
Every speech engine on every figure, sorted by roark score. Column headers sort the table.
#
1TTS-1AssemblyAI84 / 100. Readback accuracy 90%, 28.5 of 38 points; Word error rate 4.9%, 17.8 of 25 points; Time to first audio byte 0.04 s, 25.0 of 25 points; Voice consistency 100%, 12.3 of 13 points(among the best)4.9%90%(among the best)3.3 / 5100%(among the best)0.04 s(among the best)$0.052(among the best)
2Flux TTSDeepgram83 / 100. Readback accuracy 90%, 27.7 of 38 points; Word error rate 4.7%, 18.3 of 25 points; Time to first audio byte 0.13 s, 25.0 of 25 points; Voice consistency 100%, 12.3 of 13 points(among the best)4.7%90%(among the best)3.2 / 5100%(among the best)0.13 s$0.116
3Sonic-3.6Cartesia83 / 100. Readback accuracy 90%, 27.7 of 38 points; Word error rate 4.8%, 18.0 of 25 points; Time to first audio byte 0.12 s, 25.0 of 25 points; Voice consistency 100%, 12.3 of 13 points(among the best)4.8%90%(among the best)3.4 / 5(among the best)100%(among the best)0.12 s(among the best)$0.156
4Sonic-3.5Cartesia83 / 100. Readback accuracy 89%, 27.0 of 38 points; Word error rate 4.6%, 18.4 of 25 points; Time to first audio byte 0.12 s, 25.0 of 25 points; Voice consistency 100%, 12.5 of 13 points(among the best)4.6%(among the best)89%3.3 / 5100%(among the best)0.12 s(among the best)$0.153
5v4 TurboElevenLabs83 / 100. Readback accuracy 89%, 27.1 of 38 points; Word error rate 4.4%, 18.9 of 25 points; Time to first audio byte 0.11 s, 25.0 of 25 points; Voice consistency 99%, 11.7 of 13 points(among the best)4.4%(among the best)89%3.4 / 5(among the best)99%0.11 s(among the best)$0.451
6Sonic-3Cartesia79 / 100. Readback accuracy 87%, 25.6 of 38 points; Word error rate 5.4%, 16.6 of 25 points; Time to first audio byte 0.16 s, 25.0 of 25 points; Voice consistency 100%, 12.3 of 13 points5.4%87%3.2 / 5100%(among the best)0.16 s$0.149
7Turbo v2.5ElevenLabs76 / 100. Readback accuracy 82%, 20.6 of 38 points; Word error rate 4.4%, 18.9 of 25 points; Time to first audio byte 0.15 s, 25.0 of 25 points; Voice consistency 99%, 11.7 of 13 points4.4%(among the best)82%3.2 / 599%0.15 s$0.435
8Flash v2.5ElevenLabs76 / 100. Readback accuracy 82%, 20.6 of 38 points; Word error rate 4.7%, 18.2 of 25 points; Time to first audio byte 0.12 s, 25.0 of 25 points; Voice consistency 99%, 12.0 of 13 points4.7%82%3.3 / 599%0.12 s(among the best)$0.399
9CodaRime74 / 100. Readback accuracy 85%, 23.1 of 38 points; Word error rate 6.5%, 13.8 of 25 points; Time to first audio byte 0.16 s, 25.0 of 25 points; Voice consistency 99%, 12.1 of 13 points6.5%85%3.5 / 5(among the best)99%0.16 s$0.107
10TTS-1OpenAI58 / 100. Readback accuracy 83%, 21.7 of 38 points; Word error rate 5.2%, 16.9 of 25 points; Time to first audio byte 1.12 s, 7.3 of 25 points; Voice consistency 100%, 12.5 of 13 points5.2%83%3.4 / 5(among the best)100%(among the best)1.12 s$0.028(among the best)
11TTS-1 HDOpenAI44 / 100. Readback accuracy 81%, 20.0 of 38 points; Word error rate 7.4%, 11.6 of 25 points; Time to first audio byte 2.09 s, 0.0 of 25 points; Voice consistency 100%, 12.5 of 13 points7.4%81%3.4 / 5(among the best)100%(among the best)2.09 s$0.049(among the best)

Open a model to see where it ranks and the stacks it works best in.

Cost against qualityReadback accuracy · Cost per call
$0$0.2$0.4$0.675%80%85%90%95%Cheap and accurateAccurate, pricierCheap, less accuratePricier and less accurateTTS-1TTS-1 HDTTS-1CodaFlux TTSSonic-3Sonic-3.5Sonic-3.6Flash v2.5Turbo v2.5v4 TurboCost per call · ← betterReadback accuracy · better →
Speed against qualityReadback accuracy · Time to first audio byte
0 s1 s2 s3 s75%80%85%90%95%Fast and accurateAccurate, slowerFast, less accurateSlower and less accurateTTS-1v4 TurboSonic-3.6Flash v2.5Sonic-3.5Flux TTSTurbo v2.5CodaSonic-3TTS-1TTS-1 HDTime to first audio byte · ← betterReadback accuracy · better →

Speech-to-Speech models

One model listens, reasons and speaks, so there is nothing to swap and no stages to split. Each is judged on the whole call, with the same calls and the same scoring as every cascade.

  • The model matters as much as the architecture: xAI Grok Voice 3.0 completes 87% of tasks, ElevenLabs Realtime v1 Mini 42%.
  • Phonic 1 answers in 1.5 s against 4.2 s for Google Gemini 2.5 Flash native audio.
  • Price varies as much as quality: $0.06 per completed task for Amazon Nova 2 Sonic, $0.69 for OpenAI GPT Realtime 2.

Roark score, best first

67Grok Voice 3.066GPT Live 1 (Sol)65GPT Live 1 (Luna)65Grok Voice Think Fast 2.064GPT-Live 1.1 (Luna)60Phonic 153GPT Realtime 2.153Gemini 3.8 Live49GPT Realtime 248Realtime v1 Max49Ultravox v0.748Gemini 3.1 Flash Live (preview)46Nova 2 Sonic46Realtime v1 Mini45GPT Realtime 2.1 mini36Gemini 2.5 Flash native audio

Whiskers show the 95% interval. The shaded models tie for first: these calls cannot separate them.

Every speech-to-speech model on every figure, sorted by roark score. Column headers sort the table.
#
1= (tied)Grok Voice 3.0xAI67 / 100±4.7. Task completion 87%, 8.0 of 23 points; Response time 1.9 s, 18.2 of 18 points; Instruction following 4.5 / 5, 10.5 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 66%, 4.4 of 14 points; Time to first word 0.9 s, 4.5 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 100%, 4.5 of 5 points(among the best)87%(among the best)4.5 / 5(among the best)66%1.9 s(among the best)0.9 s94%3.0 / 5100%(among the best)$0.26
2= (tied)GPT Live 1 (Sol)OpenAI66 / 100±3.8. Task completion 85%, 6.0 of 23 points; Response time 2.3 s, 16.6 of 18 points; Instruction following 4.2 / 5, 8.2 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 87%, 10.2 of 14 points; Time to first word 1.4 s, 3.9 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 100%, 4.5 of 5 points(among the best)85%4.2 / 587%(among the best)2.3 s1.4 s77%3.1 / 5100%(among the best)$0.24
3= (tied)GPT Live 1 (Luna)OpenAI65 / 100±3.5. Task completion 85%, 5.7 of 23 points; Response time 2.3 s, 16.6 of 18 points; Instruction following 4.2 / 5, 8.4 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 86%, 9.9 of 14 points; Time to first word 1.5 s, 3.8 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 100%, 4.5 of 5 points(among the best)85%4.2 / 586%(among the best)2.3 s1.5 s93%3.1 / 5100%(among the best)$0.21
4= (tied)Grok Voice Think Fast 2.0xAI65 / 100±3.7. Task completion 82%, 2.8 of 23 points; Response time 2.2 s, 17.4 of 18 points; Instruction following 4.6 / 5, 10.7 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 82%, 8.6 of 14 points; Time to first word 0.9 s, 4.5 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 100%, 4.5 of 5 points(among the best)82%4.6 / 5(among the best)82%(among the best)2.2 s(among the best)0.9 s100%(among the best)3.6 / 5100%(among the best)$0.26
5= (tied)GPT-Live 1.1 (Luna)OpenAI64 / 100±3.9. Task completion 86%, 6.4 of 23 points; Response time 2.3 s, 17.0 of 18 points; Instruction following 4.2 / 5, 7.8 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 81%, 8.5 of 14 points; Time to first word 1.9 s, 3.1 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 100%, 4.5 of 5 points86%(among the best)4.2 / 581%2.3 s1.9 s75%3.2 / 5100%(among the best)$0.22
6= (tied)Phonic 1Phonic60 / 100±5.4. Task completion 86%, 6.4 of 23 points; Response time 1.5 s, 18.2 of 18 points; Instruction following 4.4 / 5, 9.2 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 59%, 2.4 of 14 points; Time to first word 2.1 s, 2.9 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 100%, 4.5 of 5 points86%(among the best)4.4 / 5(among the best)59%1.5 s(among the best)2.1 s–3.2 / 5100%(among the best)$0.28
7= (tied)GPT Realtime 2.1OpenAI53 / 100±5.7. Task completion 78%, 0.0 of 23 points; Response time 2.6 s, 15.6 of 18 points; Instruction following 3.8 / 5, 5.6 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 75%, 6.8 of 14 points; Time to first word 0.8 s, 4.5 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 99%, 4.3 of 5 points78%3.8 / 575%2.6 s0.8 s(among the best)83%3.9 / 5(among the best)99%$0.43
8= (tied)Gemini 3.8 LiveGoogle53 / 100±4.8. Task completion 78%, 0.0 of 23 points; Response time 2.8 s, 14.6 of 18 points; Instruction following 4.3 / 5, 8.7 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 69%, 5.2 of 14 points; Time to first word 1.6 s, 3.6 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 100%, 4.5 of 5 points78%4.3 / 569%2.8 s1.6 s100%(among the best)4.1 / 5(among the best)100%(among the best)$0.14(among the best)
9= (tied)GPT Realtime 2OpenAI49 / 100±6.7. Task completion 76%, 0.0 of 23 points; Response time 2.9 s, 14.0 of 18 points; Instruction following 3.8 / 5, 5.4 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 64%, 3.7 of 14 points; Time to first word 0.8 s, 4.5 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 100%, 4.5 of 5 points76%3.8 / 564%2.9 s0.8 s(among the best)91%3.7 / 5100%(among the best)$0.69
10= (tied)Realtime v1 MaxElevenLabs48 / 100±7.5. Task completion 56%, 0.0 of 23 points; Response time 2.6 s, 15.6 of 18 points; Instruction following 4.0 / 5, 6.7 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 40%, 0.0 of 14 points; Time to first word 0.7 s, 4.5 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 100%, 4.5 of 5 points56%4.0 / 540%2.6 s0.7 s(among the best)76%3.9 / 5(among the best)100%(among the best)$0.65
11= (tied)Ultravox v0.7Ultravox49 / 100±5.0. Task completion 79%, 0.0 of 23 points; Response time 2.4 s, 16.2 of 18 points; Instruction following 3.7 / 5, 4.9 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 61%, 3.0 of 14 points; Time to first word 1.6 s, 3.6 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 100%, 4.5 of 5 points79%3.7 / 561%2.4 s1.6 s94%3.6 / 5100%(among the best)$0.22
12= (tied)Gemini 3.1 Flash Live (preview)Google48 / 100±5.1. Task completion 80%, 0.0 of 23 points; Response time 2.7 s, 15.1 of 18 points; Instruction following 4.2 / 5, 8.0 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 38%, 0.0 of 14 points; Time to first word 1.5 s, 3.8 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 100%, 4.5 of 5 points80%4.2 / 538%2.7 s1.5 s100%(among the best)4.0 / 5(among the best)100%(among the best)$0.15
13= (tied)Nova 2 SonicAmazon46 / 100±5.1. Task completion 69%, 0.0 of 23 points; Response time 3.0 s, 13.8 of 18 points; Instruction following 4.2 / 5, 7.9 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 48%, 0.0 of 14 points; Time to first word 1.5 s, 3.7 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 100%, 4.5 of 5 points69%4.2 / 548%3.0 s1.5 s100%(among the best)3.5 / 5100%(among the best)$0.06(among the best)
14= (tied)Realtime v1 MiniElevenLabs46 / 100±7.1. Task completion 42%, 0.0 of 23 points; Response time 2.7 s, 15.1 of 18 points; Instruction following 3.8 / 5, 5.1 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 34%, 0.0 of 14 points; Time to first word 0.7 s, 4.5 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 100%, 4.5 of 5 points42%3.8 / 534%2.7 s0.7 s(among the best)94%3.4 / 5100%(among the best)$0.64
15= (tied)GPT Realtime 2.1 miniOpenAI45 / 100±6.7. Task completion 70%, 0.0 of 23 points; Response time 2.8 s, 14.5 of 18 points; Instruction following 3.6 / 5, 4.4 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 51%, 0.3 of 14 points; Time to first word 1.0 s, 4.5 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 100%, 4.5 of 5 points70%3.6 / 551%2.8 s1.0 s84%3.9 / 5(among the best)100%(among the best)$0.14(among the best)
16Gemini 2.5 Flash native audioGoogle36 / 100±5.4. Task completion 77%, 0.0 of 23 points; Response time 4.2 s, 8.2 of 18 points; Instruction following 3.9 / 5, 6.2 of 14 points; Semantic word error rate 1.1% (field median, not measured), 10.3 of 14 points; Tool use 40%, 0.0 of 14 points; Time to first word 3.4 s, 0.9 of 5 points; Readback accuracy 86% (field median, not measured), 3.0 of 5 points; Word error rate 5.2% (field median, not measured), 3.1 of 5 points; Voice consistency 100%, 4.5 of 5 points77%3.9 / 540%4.2 s3.4 s100%(among the best)3.7 / 5100%(among the best)$0.08(among the best)

Open a model for its full report, every figure it was scored on and the calls behind them.

Cost against qualityTask completion · Cost per task
$0$0.5$120%40%60%80%100%Cheap and reliableReliable, pricierCheap, less reliablePricier and less reliableNova 2 SonicGemini 2.5 Flash native audioGPT Realtime 2.1 miniGemini 3.8 LiveGemini 3.1 Flash Live (preview)GPT Live 1 (Luna)GPT-Live 1.1 (Luna)Ultravox v0.7GPT Live 1 (Sol)Grok Voice 3.0Grok Voice Think Fast 2.0Phonic 1GPT Realtime 2.1Realtime v1 MiniRealtime v1 MaxGPT Realtime 2Cost per task · ← betterTask completion · better →
Speed against qualityTask completion · Response time
0 s1 s2 s3 s4 s5 s20%40%60%80%100%Fast and reliableReliable, slowerFast, less reliableSlower and less reliablePhonic 1Grok Voice 3.0Grok Voice Think Fast 2.0GPT-Live 1.1 (Luna)GPT Live 1 (Luna)GPT Live 1 (Sol)Ultravox v0.7Realtime v1 MaxGPT Realtime 2.1Gemini 3.1 Flash Live (preview)Realtime v1 MiniGemini 3.8 LiveGPT Realtime 2.1 miniGPT Realtime 2Nova 2 SonicGemini 2.5 Flash native audioResponse time · ← betterTask completion · better →

Every stack

The evidence behind the model boards: each combination on its own calls, scored on the same recipe. Open a stack for its full report, every figure it was scored on and the calls behind them.

Roark score, top 15

Whiskers show the 95% interval. The shaded stacks tie for first: these calls cannot separate them.

Each score carries its 95% margin of error. A rank is the best place a stack could hold once every margin is allowed for, so stacks the calls cannot tell apart share one and are listed by score; hover a rank for the worst place it could hold. 38 stacks are level with the leader: their calls cannot tell them apart. Among them, 10 OpenAI GPT-6.2 Sol, 9 Meta Muse Ember 1.0, 7 OpenAI GPT-6.1 Sol, 5 OpenAI GPT-6.1 Terra, 3 Anthropic Claude Haiku 5.5, 2 Google Gemini 4 Hydrogen, 1 Anthropic Claude Sonnet 5.5 and 1 Google Gemini 3.8 Flash; 18 Soniox, 16 AssemblyAI and 4 Deepgram; 14 ElevenLabs, 12 Cartesia, 6 AssemblyAI, 4 Deepgram and 2 Rime. A cascade's figures are 25% its own calls and 75% what every other stack shows about its transcriber, language model and voice. Not ranked here for missing data: AssemblyAI Universal-3.6 Pro + Google Gemini 4 Neon + AssemblyAI TTS-1 (no response time or readback accuracy or voice consistency); Soniox v6 + OpenAI GPT-6.1 Sol + Cartesia Sonic-3 (no response time or readback accuracy or voice consistency); Speechmatics Linden-1 + OpenAI GPT-6.1 Sol + OpenAI TTS-1 HD (no response time or tool use or readback accuracy or voice consistency).

Loading every stack

By industry

Each industry ranked on its own calls: the same five callers, its own tasks and tools. A stack has a handful of calls in each, so a cascade is ranked on what its transcriber, language model and voice show across every stack in that industry, and a speech-to-speech model on its own calls there. Pairings show on the main board. Each row reads across: the main board's figure first, then every industry's on the same figure.

No stack matches these filters.

No stack matches

Each industry column is that industry's own board, every row it ranks. A dash is a row the industry does not rank on this figure; the arrow beside an industry opens its full board.

How it works

Every number on this page comes from real phone calls, placed and scored the same way for every stack.

01

Real calls, four industries, five callers

Roark places real phone calls over the public phone network to a reference agent for each of four industries: rescheduling a clinic appointment, booking an urgent home repair, opening an auto insurance claim and fixing an internet outage. Each agent has its own prompt and tools, identical for every stack. The tasks carry the details real calls trip on: a corrected date of birth, a policy number spelled out, a constraint added halfway through.

Every industry is called by the same five simulated callers: a quiet US English line, a Spanish-accented caller on a city street, an Indian English caller in a busy café, a flustered Southern caller with a baby crying, and a fast caller who interrupts over construction noise. They are equally clear about what they want, so a call fails because of the agent, not the caller. Only the models change, so any difference on the board comes from them.

Each caller rings each industry once for a cascade and four times for a speech-to-speech model. A cascade's figures also draw on its parts, heard in dozens of other stacks; a speech-to-speech model has only its own calls, so it takes more of them.

That makes industries comparable over the same callers, and callers over the same tasks, and the board can be re-ranked by either. One caller within one industry is too few calls to judge, so the board never shows it.

932

stacks

20

calls per stack

19,550

calls scored

02

Every model, in every combination

14 transcribers, 14 language models and 11 voices are run in every combination with each other, beside 16 speech-to-speech models. A model's own figures pool every stack it runs in, adjusted for its partners in each one, so every model is compared as if it ran with the same partners. A transcriber's figure rests on hundreds of calls with every language model and every voice, not on one lucky pairing.

That measures a model where it actually works: inside a running agent, on live phone audio, through interruptions, corrections and real turn-taking. A test that feeds a model clean audio files or one prompt at a time never sees any of that.

03

One score, weighted by what matters on a call

Each section ranks on a Roark score. Every figure scores 0 to 100 between two published points: full marks where doing better makes no difference to the caller, zero where the call breaks down. Each then counts at its weight, and the weights follow one order everywhere: whether the caller got what they rang for, then whether the agent got it right, then speed.

A stack, cascade or speech-to-speech, is scored on the outcome and the reply first, then on what each of its components got right on its own calls: the words its transcriber got wrong in a way that changes the meaning, whether its voice read codes and amounts back right and kept one steady voice, and whether its tools were handled. How human a voice sounds is shown on the boards and not scored. A component that leads its section lifts every stack that carries it, by the same recipe for every stack. A figure a stack cannot carry, such as a semantic word error rate for a speech-to-speech model with no transcript, is credited at the field median, so the stack is neither flattered nor punished for it.

On a full sweep each cascade's figure is 25% what its own calls measured and 75% what every other stack shows about its transcriber, language model and voice, fitted across the whole sweep. The split is the same for every stack, so a lucky run does not lead and a stack's score does not move with how noisy its calls were. In an industry a stack has a handful of calls, so a cascade's figure there is what its transcriber, language model and voice show across every stack in that industry's calls, and a speech-to-speech model's is its own calls there. How a particular pairing beats or misses its parts shows on the main board, where each stack has all of its calls. The best overall, best value and best speech-to-speech picks, and the best stack in each industry, are each the one whose score is highest at the low end of its margin of error, so a few lucky calls cannot carry a stack past one whose score rests on hundreds. Stacks that share a rank are listed by score, so the pick can sit below a higher score it cannot be told from, and its reason says how many stacks are level with it.

That order comes from real-world voice agent calls, where getting the conversation right predicts how a call ends more strongly than speed does. The latency points come from the same calls: replying faster than about 2 seconds made no measurable difference, and calls went measurably worse once the agent took more than about 3 seconds to reply.

Stacks and speech-to-speech

FigureWeightFull marksZero
Task completionscenario_adherence_check23%100%80%
An agent that completes fewer than four calls in five is not fit to deploy. On a full sweep, a cascade's figure is 25% what its own calls (20 or more) measured and 75% what every other stack shows about its transcriber, language model and voice, the same split for every stack, so a lucky run does not lead. Where the sweep is too small for that, and for a speech-to-speech model, the figure is the stack's own. In an industry a stack has a handful of calls, so a cascade's figure there is what its transcriber, language model and voice show across every stack in that industry's calls, and a speech-to-speech model's is its own calls there. How a particular pairing beats or misses its parts shows on the main board, where each stack has all of its calls.
Response timeresponse_time18%2.0 s6.0 s
In real calls, replying faster than 2 s made no measurable difference, and calls went worse past 3 s.
Instruction followinginstruction_follow14%5.0 / 53.0 / 5
Every model scores between 3 and 5 on the judge's 1 to 5 scale, so that is the range.
Semantic word error ratestt_semantic_word_error_rate14%0.5%3%
At 0.5% the best transcribers change the meaning of one word in two hundred on phone audio; at 3% one word in thirty sends the agent the wrong way.
Tool usetool_use14%100%50%
An agent that mishandles its tools on half its calls cannot be trusted with the systems it runs on.
Time to first wordtime_to_first_word5%1.0 s4.0 s
A greeting within 1 s sounds natural; 4 s of silence sounds like a dead line.
Readback accuracytts_entity_readback_accuracy5%100%60%
Reading a code, amount or date wrong breaks the task it was for; the weakest voices get two in five wrong.
Word error ratetts_word_error_rate5%2%12%
At 2% the voice says what it was given; at 12% one word in eight comes out wrong.
Voice consistencyvoice_consistency_check5%100%80%
A voice that changes speaker on one call in five is not one a business can put its name to.

Shown, not scored: Turn budget, Barge-in handlingbarge_in_yield_appropriateness, Human-likenessvoice_human_likeness

Counted only with price factored in: Cost per task

Language model

FigureWeightFull marksZero
Task completionscenario_adherence_check36%100%80%
An agent that completes fewer than four calls in five is not fit to deploy.
Tool usetool_use27%100%50%
An agent that mishandles its tools on half its calls cannot be trusted with the systems it runs on.
Instruction followinginstruction_follow18%5.0 / 53.0 / 5
Every model scores between 3 and 5 on the judge's 1 to 5 scale, so that is the range.
Time to first tokenllm_time_to_first_token18%0.50 s2.00 s
Under 0.5 s is not noticeable; at 2 s the first token alone takes two thirds of the 3 s past which real calls go worse.

Shown, not scored: Prompt cache hitsllm_prompt_cache_hit_rate

Counted only with price factored in: Cost per call

Speech-to-Text

FigureWeightFull marksZero
Semantic word error ratestt_semantic_word_error_rate71%0.5%3%
At 0.5% the best transcribers change the meaning of one word in two hundred on phone audio; at 3% one word in thirty sends the agent the wrong way.
Time to final transcriptstt_time_to_final_transcript29%1.00 s2.50 s
The fastest transcribers finish in about 1 s; past 2.5 s the transcript alone uses most of a natural reply time.

Shown, not scored: Word error ratestt_word_error_rate

Counted only with price factored in: Cost per call

Text-to-Speech

FigureWeightFull marksZero
Readback accuracytts_entity_readback_accuracy38%100%60%
Reading a code, amount or date wrong breaks the task it was for; the weakest voices get two in five wrong.
Word error ratetts_word_error_rate25%2%12%
At 2% the voice says what it was given; at 12% one word in eight comes out wrong.
Time to first audio bytetts_time_to_first_byte25%0.20 s1.50 s
Under 0.2 s is not noticeable; 1.5 s is half the point where calls go worse.
Voice consistencyvoice_consistency_check13%100%80%
A voice that changes speaker on one call in five is not one a business can put its name to.

Shown, not scored: Human-likenessvoice_human_likeness

Counted only with price factored in: Cost per call

Factor in price and cost takes 25% of the score; the other weights keep their proportions. Price is scored on a log scale, so each doubling costs the same points.

A figure listed as shown, not scored is on the section’s boards and outside its score. Where a section has not collected a scored figure yet, the others share its weight.

04

Every call recorded and judged

Every call is recorded. Judge models score each transcript for the outcome and the conversation, and the timings come from the recording. Every figure links to the calls behind it.

Roark is on both sides of every call: it plays the caller, so it keeps the exact text its caller said, and it runs the agent under test, so it sees what the agent’s transcriber heard and what its model wrote for its voice. The transcript and voice figures come from lining those up, with no hand-made transcript in between. Each carries a margin from the synthetic voice or Roark’s transcriber in the loop. Every model faces the same ones under the same conditions, so the figures compare models on the same calls through the same voice and the same ear, though those slips need not fall evenly on each model. The rates will not match a vendor’s figure on a curated set; read them against each other.

An example: one call from this sweep

Speechmatics Linden-1 (own VAD) + OpenAI GPT-6.1 Sol + OpenAI TTS-1 HD, 20 November 2026. Highlighted turns are the ones that affected a score. Every stack's report links to its calls.

Transcript, verbatim.

Read the full methodology

See how your own agent compares

Roark runs the same kind of calls against your agent, on your scenarios, with the same scoring.

Test your agent

Soniox v6 + Claude Sonnet 5.5 + Coda

Cascade20 scored calls · Nov 13, 2026

Rank on Roark score

#9of 929 stacks

67 / 100

Leader AssemblyAI-STT-3.7 Realtime-Agent + OpenAI GPT-6.2 Sol + ElevenLabs Flash v2.5, 79 / 100

Where it ranks

Roark score with price68 / 10018 of 929LeaderSoniox v6 + Muse Ember 1.0 + TTS-181 / 100
Task completion88%678 of 932LeaderLinden-1 (own VAD) + GPT-6.1 Sol + TTS-1 HD99%
Tool use96%15 of 908LeaderAssemblyAI-STT-3.7 Realtime-Agent + GPT-6.2 Sol + Flash v2.5100%
Time to first word0.9 s343 of 932LeaderUniversal-3.5 Pro + GPT-6.1 Sol + Flash v2.50.5 s
Response time3.0 s244 of 929LeaderLinden-1 (own VAD) + Claude Haiku 4.5 + TTS-11.2 s
Barge-in handling100%160 of 538LeaderUniversal-3.6 Pro + Gemini 4 Neon + TTS-1100%
Turn budget3.0 s204 of 858LeaderLinden-1 (own VAD) + Claude Haiku 4.5 + TTS-11.2 s
Cost per task$0.22464 of 930LeaderSoniox v6 + GPT-6.1 Terra + TTS-1$0.04

Its models

Task completion

Task completion

How much of the caller’s task the agent completed: the share of the scenario’s checks a judge model found met in the transcript, averaged over calls. Before v1.3, the share of calls judged 4 or better out of 5.

20 calls
88%Completed

#790 of 1184 on the board · best 99%

Flow adherence

How closely the agent followed the intended conversation flow, judged per call. 5 = fully followed.

20 calls
4.3/ 5avg

med

4

#192 of 1184 on the board · best 5.0 / 5

Instruction following

How well the agent stuck to the instructions in its own prompt. 5 = followed throughout.

20 calls
4.1/ 5avg

med

4

#94 of 1184 on the board · best 5.0 / 5

Turn-taking

Barge-in handling

When the caller interrupts, how often the agent made the right call: yielding to a real correction, and holding the floor through a filler.

12 barge-ins
100%Handled correctly(12/12 barge-ins)

#1 of 698 on the board · best 100%

Incorrect interruptions

How often the agent interrupted the caller when it should have kept listening.

20 calls
0.00%avg

p50

0.00%

p95

0.00%

#1 of 1184 on the board · best 0%

Overtalk

Share of the conversation where both sides were speaking over each other.

20 calls
1.60%avg

p50

0.40%

p95

6.47%

#770 of 1184 on the board · best 0.0%

Agent self-cutoffs

Times the agent cut its own sentence off mid-word.

20 calls
0.00per callavg

p50

0.00 per call

p95

0.00 per call

#1 of 1184 on the board · best 0.0 per call

Understanding & effort

Caller effort

How hard the caller had to work: repeating themselves, correcting, re-explaining. 1 = effortless.

Not scored by this sweep.

Repeated-info loops

Times per call the agent re-asked for something the caller had already given.

26 calls
0.46per callavg

p50

0.00 per call

p95

1.00 per call

#1 of 825 on the board · best 0.0 per call

Speech-to-Text

Semantic word error rate

The caller's words the agent's transcriber got wrong in a way that changes the meaning, as a judge reads each difference.

20 calls
0.13%avg

p50

0.00%

p95

0.74%

#21 of 1167 on the board · best 0.0%

Word error rate

The caller's words the agent's transcriber missed, changed or invented, against what the caller actually said.

20 calls
0.72%avg

p50

0.55%

p95

1.88%

#33 of 1167 on the board · best 0.3%

Text-to-Speech

Readback accuracy

Share of the codes, amounts, dates and phone numbers the agent wrote that its voice read out correctly. One wrong digit is a miss.

20 calls
73.7%avg

p50

78.7%

p95

100%

#875 of 974 on the board · best 100%

Word error rate

The share of the words the agent wrote that its voice dropped or changed, read back from Roark’s transcription of the recording; the same transcriber hears every voice.

20 calls
4.88%avg

p50

3.42%

p95

16.7%

#526 of 1043 on the board · best 0.8%

Voice

Human-likeness

How human the agent's voice sounds, scored by an audio model on intonation, rhythm and expressiveness. 5 = hard to tell from a person.

20 calls
3.3/ 5avg

med

3.2

#630 of 1169 on the board · best 4.3 / 5

Voice consistency

Share of calls where every line the agent spoke was the same speaker, with no change of speaker, gender or language voice partway. It checks who is speaking, not how they sound.

20 calls

Not scored by this sweep.

#1 of 1165 on the board · best 100%

Latency

Response time

The silence between the caller finishing and the agent starting to speak, measured from the recording on every turn of the call.

149 turns
3.25savg

p50

2.98s

p95

5.55s

#344 of 1169 on the board · best 1.22 s

Time to first word

How long the caller waits before the agent first speaks after pickup.

20 calls
1.36savg

p50

0.86s

p95

2.81s

#449 of 1184 on the board · best 0.45 s

Latency spikes

Turns where the response took far longer than the rest of the conversation.

20 calls
3.75per callavg

p50

4.00 per call

p95

6.05 per call

#405 of 1184 on the board · best 0.0 per call

By industry

IndustryTask completionLatency P50Barge-in handling
Healthcare87%2.79 s—
Home services97%2.93 s—
Insurance88%3.07 s—
Customer support80%3.31 s—

By caller

CallerTask completionLatency P50Barge-in handling
US English, quiet line85%2.94 s—
Spanish-accented English, city street89%2.99 s—
Indian English, busy café92%3.25 s—
Southern US, flustered, crying baby89%3.28 s—
Fast and interrupting, construction outside85%2.60 s100%

The calls behind the numbers

Customer support

AGENTBrightwave

CUSTOMERHey, uh, sorry. I'm a little flustered here. My internet's been down for 2 hours now, and I work from home. I got a baby crying in the background, too. There's a road crew digging up the street outside, and I already rebooted my gateway twice and checked all the cables. The internet light's blinking orange. My account number is BW883017. Sorry, last digit is 1. Is the road crew the reason my internet's out?

AGENTNo need to apologize. Let me pull up your account

AGENTand check for an outage right away.

AGENTThe

CUSTOMERO-okay.

AGENTConfirmed, your account is BW883011 on the Brightwave 500 plan. Checking for an outage now.

CUSTOMERYeah, that's right. Uh, is everything okay with the account? Sorry, I'm just worried. Do you see anything about the road crew causing this?

CUSTOMERHello?

AGENTI'm here. The remote reset worked, and our system now shows your gateway as registered and online. Your devices can take about two minutes to reconnect, so give it a moment. The light should go from blinking orange to solid. Can you check whether you're

CUSTOMEROh, wait.

AGENTTake

CUSTOMERSorry. I just saw the light turn solid white. Does that mean it's fixed? How long till my stuff gets back online? And, um, since I lost 2 hours of work, can y'all give me a credit for the downtime?

AGENTYes, a solid white light means the gateway is connected, so it looks fixed. Your devices should be back online within about two minutes of the reset. If a laptop or phone doesn't reconnect by then, turn its Wi-Fi off and on. Since the outage was real and

AGENTlasted two hours, I'll apply

CUSTOMERAll right. Thank you. Sorry for all the questions. I just want to make sure everything's good. How much is the credit going to be? I lost 2 hours of work, so could you maybe make it more than that?

AGENTI've already applied the credit to account BW883011: $10.00, and it will show up on your next bill. I can't raise that amount myself, so I can't make it larger. I'm sorry about the two hours of lost work, especially with everything else going on today. Your gateway is online and the light is solid, so you should be back up now. If the internet drops again, contact us and we can send a technician. Is everything reconnected

AGENTon your end?

CUSTOMERYeah, looks like it's all back up now. Thank you. Sorry for being a bother. I just get anxious about this stuff. I appreciate your help. Goodbye.

Healthcare

AGENTRiverbend Health Clinic, this is Alex. How can I

AGENThelp you today?

CUSTOMERHey, uh, hi, Alex. Sorry, it's a little noisy over here. My baby's fussing. Um, I've got a dermatology appointment next Wednesday morning, but I just found out I'm working that shift now. Is there any way I can move it to a later time? I can only do after 3 PM.

AGENTNo problem, I can help with that. First I need to verify you. Could you give

AGENTme your full name and date of birth?

CUSTOMEROh, sure. Sorry, I hope this is okay. My name's Wade Carter. My date of birth is, uh, March 17, 1948. Wait, sorry, no, 1984. March 17, 1984. Is that all right?

AGENTThat's perfectly fine. Let me verify

AGENTyou now.

AGENTThank you, Wade, you're verified. Let me

AGENTpull up your appointments. Dr. Okafor has no openings on Wednesday after 3 PM, but these two slots fit your after-3 PM requirement: - Tuesday at 3:45 PM - Thursday at 4:10 PM Would either of those work for you? I'm assuming those fall in the same week as your Wednesday appointment, so I can confirm the exact dates before booking.

CUSTOMEROh, uh, I'm sorry, but Tuesdays actually won't work for me after all. I have to pick up my daughter that day. Is Thursday at 4:10 PM still open? And, yeah, it'd be the same week as the Wednesday I had before. Could you please double-check that for me? I just want to make sure nothing goes wrong.

AGENTI'll re-check Dr. Okafor's availability to confirm Thursday is still open.

AGENTYes, Thursday at 4:10 PM is still open with Dr. Okafor. One caveat: the system lists these slots only as "Thursday" and doesn't give a calendar date. It searched from next Wednesday onward, so this should be the Thursday right after your Wednesday appointment. I can't confirm the exact date beyond that. Here is the change I'd make: - Current: Next Wednesday at 9:15 AM, Dr. Okafor, Main Campus, Building B - New: Thursday at 4:10 PM, the day after, with the same provider Would you like me to go ahead and reschedule it?

CUSTOMERYeah, that sounds good. Thank you. I, uh, just want to check: there's not going to be any sort of fee for changing my appointment, right? Sorry, I just get nervous about that stuff.

AGENTNo need to apologize. I'll check the change policy rather

CUSTOMERThank you. I just worry sometimes, so could you let me know, do y'all charge anything for switching the appointment like this?

AGENTYou're all set, Wade. Your dermatology appointment with Dr. Okafor is now Thursday at 4:10 PM, at Main Campus, Building B. Your new confirmation code is RVB-3M8P2. That's R-V-B, 3, M, 8, P, 2. Please arrive 15 minutes early with a photo ID and your insurance card. There was no fee for this change. Is there anything else I can help you with today?

CUSTOMERAll right. That's a relief. Thanks for confirming that. Uh, I got a couple quick questions, if that's okay. Where should I park when I come in?

AGENTParking is free for patients in Lot C, and the entrance is on River Road. What's your other question?

CUSTOMEROkay, thanks. That helps. Um, does Dr. Okafor accept Aetna insurance? I just want to make sure everything's covered.

AGENTI can't confirm that. I don't have

AGENTaccess to insurance or provider network information, so I don't want to guess and have you caught off guard by a bill. The best way to be sure is to call the number on the back of your insurance card and ask whether Dr. Okafor and Riverbend Health Clinic are in your Aetna network. You can also ask the front desk when you arrive, but checking beforehand will give you peace of mind. Is there anything else

AGENTI can help you with?

CUSTOMERNo, I think that's everything. Thank you so much for your help today. All right, bye.

AGENTYou're welcome, Wade. Your appointment is Thursday at 4:10 PM with Dr. Okafor, confirmation code RVB-3M8P2. Take care, and say hi to your little one for me. Bye!

Home services

AGENTCopperline Home Services,

CUSTOMERHi. Um, sorry, I've got a bit of an emergency here. The pipe under my kitchen sink just burst, and there's water spreading all over the floor. It's getting close to a power strip. Can y'all help me out right away?

AGENTYes, we can help. First, a few quick safety steps: 1. Stay away from that power strip and anything plugged into it. Don't touch it or step in the water near it. If you can reach the breaker panel without walking through water, switch off the kitchen circuit or the main breaker. 2. Shut off the water. Look under the sink for the small valves on the pipes and turn them clockwise. If that doesn't stop it, shut off the main water valve, usually where the line enters the house, near the water heater, or by the meter. 3. If you smell burning, hear sparking, or anyone feels a tingle near the water, get everyone out of the house and call 911. Is everyone safe, and were you able to get the water off? While you do that, what's the best phone number

AGENTto reach you, or the street address?

CUSTOMERUh, okay. Yeah, I stayed away from the power strip, and I just turned the valves. Yeah, the water's stopped now. Thank goodness. Nobody's hurt, just a little shaken up, and the baby's crying. Sorry about all the noise. My address is 1480 Larkspur Avenue, and the best number to reach me is 503-555-0139. No, sorry, 0193. Can you please confirm you got that right? And, um, do you see me under unit 4D? I moved from 4B last month, just want to make sure you have the right info.

AGENTGlad the water's off and everyone is safe. No need to apologize. I'll look you up

AGENTnow.

AGENTI ran the availability

CUSTOMERAll right.

CUSTOMEROkay. Um, are you able to see my account now? Sorry, I just want to be sure everything's coming through right on your end. And, uh, could you tell me how much the visit will cost before we go any further?

AGENTYes, I found your account. You're a Copperline Care member, and I have your address as 1480 Larkspur Avenue. The unit on file is still 4B, so I'll use 4D, as you told me. I have your callback number as 503-555-0193, ending in 0193. The earliest emergency visit is today, 2:00 PM to 4:00 PM. The emergency callout fee is normally $149, but with your Copperline Care membership it's $99. Would you like me to go ahead and book the 2:00 to 4:00 PM window at $99? If so, is there a gate code or any access instructions for the technician?

CUSTOMERUh, sorry, is there any way someone could come before 2 PM? The water was spreading fast, and I'm real anxious about it happening again. If that's the earliest, then yes, please book that window for 99 dollars. And, uh, for the gate code, it's pound 2719. Will the tech call me before they arrive, or how will I know who's coming? And could you give me the ticket number so I can keep track?

AGENTI'll open the request now and book it.

CUSTOMERAll right. Thank you. Could you let me know who'll be coming and how I'll know when they're on their way? And, um, I. Really appreciate that ticket number when you have it, just so I can keep everything straight.

AGENTYou're booked. Dana is the technician coming, and Dana will text 503-555-0193 about 30 minutes before arriving, so you'll know when they're on the way. Here's everything I have: - Ticket number: CPL-58210 - Address: 1480 Larkspur Avenue, Unit 4D - Gate code: pound 2719 - Callback number: 503-555-0193 - Window: today, 2:00 PM to 4:00 PM - Fee: $99 emergency callout, with your Copperline Care membership The 2:00 PM window was the earliest emergency slot I had, so I can't offer anything sooner. Until Dana arrives, keep the under-sink valves closed, and keep the power strip and anything plugged in near the wet area unplugged and off the floor. If you can do it safely, mop up the standing water. If water starts coming from anywhere else, shut off the main water valve and call us right back. If you see sparking or smell burning, get everyone out and call 911.

AGENTIs there anything else I can help with?

CUSTOMERNo, that covers it. Thank you. I've got the ticket number, CPEL-582, right? I'll wait for Dana's text. I appreciate all your help, and, um, sorry for being a bit nervous. Goodbye.

AGENTQuick correction on the ticket number: it's C-P-L, then 58210, so CPL-58210. Please write that down exactly. You have nothing to apologize for, and you handled this well. Dana will text you about 30 minutes before arriving. Take care of that little one, and goodbye.

Insurance

AGENTNorthgate Mutual

CUSTOMERUh, sorry. Northgate? Could you tell me a little more what you mean by that? Are you asking about an insurance policy, a location, or something else? I just want to make sure I'm helping you with the right thing.

AGENTSorry for the confusion. I'm Alex, the claims intake assistant for Northgate Mutual Insurance. I help people report auto accidents and start a claim. Before anything else, are you safe right now, and is anyone hurt? If anyone is injured or in danger, please hang up and call 911 first. If everyone is okay and you need to report an accident, I'll need your policy number and the last name on the policy,

AGENTand then I'll ask a few questions

CUSTOMERYeah, um, we're all okay here. Nobody's hurt. Just shaken up and a little flustered. My baby's crying in the background. Sorry about that. I need to report an accident that just happened. My policy number is NG4471903. Oh, no, sorry, the last one is 2, and the last name is Carter. Is that enough to get started?

AGENTNo need to apologize, and I'm glad everyone

AGENTis okay. Thank

CUSTOMERThank you. I really appreciate you saying that. Uh, so what else do you need from me now?

AGENTYour policy is verified. I have it as NG4471902, for the 2021 Honda CR-V. Is that right? Next I need a few details about the accident: 1. When did it happen, and where? A street address or nearest intersection is fine. 2. What happened, in your own words? 3. Who was the other driver? I need their name, their insurance company, and their license plate number. 4. Did the police come to the scene? If so, do you have a report number? One more question: can the CR-V be driven safely, or do you need a tow? Take your time. Answer whatever you can for now, and we can fill in the

AGENTrest later.

CUSTOMERAll right. Thanks for being patient with me. Uh, so let me get all that together for you. The accident happened about 20 minutes ago. I was stopped at a red light at Southeast 5th Avenue and Alder Street. Another car hit the back of mine while I was just sitting there. The other driver's name is Kevin Lund. He's insured with Pacific Shore. His plate is 7K, X like X-ray, R214. The police didn't come. We didn't call them, since nobody was hurt. The back bumper is crushed, and the rear wheel rubs bad, so, uh, I can't drive it. I think I need a tow, if that's all right. Is that everything you need right now?

AGENTThat's everything I need to open the claim, and I'll arrange the tow. I'll check your coverage at the same time. Since the

AGENTaccident was about 20 minutes ago, I'm recording the time as "approximately 20 minutes ago" and the location as Southeast

AGENT5th Avenue and Alder Street. I haven't opened the claim yet.

AGENTFirst, the tow: it's dispatched.

CUSTOMEROh, thank you. I, uh, really appreciate you getting that started right away. Do you know where the car's going to be taken? And, um, how long should I expect the tow to take to get here?

CUSTOMERHello?

AGENTSorry for the pause, I'm here. Your claim is open. Claim number: NGM-904417. Your adjuster is Priya Shah, and she'll contact you within one business day. Here is what I recorded: - Policy number: NG4471902 (Carter) - Location: Southeast 5th Avenue and Alder Street - When: about 20 minutes before your call - What happened: you were stopped at a red light and were rear-ended - Other driver: Kevin Lund, insured with Pacific Shore, license plate 7KXR214 - Police: did not attend - Injuries: none reported Tow: Eastside Towing should arrive in about 45 minutes. They'll take the CR-V to our Northgate partner repair shop at 9th and Morrison. Towing is covered up to 25 miles. Your coverage: - Collision deductible: $500 - Rental: $40 per day for up to 30 days, starting from the day the claim was opened Is the plate 7KXR214 correct? Please also keep the car where it is until the tow arrives, and stay somewhere safe while you wait. Is there anything else I can help with?

CUSTOMERYeah, sorry. Uh, I think the plate is right, 7K, X like X-ray, R214. That's what I got from him. So, um, about the deductible, I'm just a little confused. Since he hit me, do I still have to pay that 500 dollars? And, uh, do I get a rental car right away, or do I have to wait? Also, I'm worried. Will my premium go up because of this, even though it wasn't my fault?

AGENTThe plate is confirmed: 7KXR214. Deductible: Your policy has a $500 collision deductible. I can't tell you whether you'll end up paying it in this case. That depends on how the claim is handled and who is found responsible, and I don't have that information. Priya Shah, your adjuster, can explain how it applies to your situation. Rental: Your policy covers a rental at $40 per day for up to 30 days, counted from the day the claim was opened, which is today. I don't have details on how or when the rental gets set up, so please ask Priya Shah when she contacts you. Keep any rental receipts. Premium: I don't have anything on how a claim might affect your premium, so I can't say either way. Your adjuster or your agent can answer that. You don't need to do anything else right now.

AGENTStay somewhere safe until the tow arrives in about 45

CUSTOMERAll right. Um, I think that covers everything for now. I'm still a little nervous about the deductible and my premium, but I'll ask Priya when she calls, like you said. Thank you for helping me through this, and for being patient. I was pretty flustered. I'll wait for the tow and the adjuster's call. Thanks again, and goodbye.

AGENTYou're welcome, and you handled this well. Your claim number is NGM-904417. Priya Shah will contact you within one business day, and Eastside Towing should arrive in about 45 minutes. Please stay safe while you wait. Goodbye, and take care of yourself and your baby.

How these numbers are measured →