CekuraBench

GPT-Live 1 + GPT-6 SolvsGrok Voice 3.0

All comparisonsLeaderboard
Compare
GPT-Live 1 + GPT-6 SolOpenAI
Wins 3 of 8
Reliability
75.6%
Response time
2.02s
Cost / min
$0.055
Grok Voice 3.0xAI
Wins 5 of 8
Reliability
81.7%
Response time
1.34s
Cost / min
$0.080

Grok Voice 3.0 is ahead on reliability, success rate, calls where the agent went quiet, interruption and response time, and GPT-Live 1 + GPT-6 Sol on agent responses, data accuracy and cost.

Last updated Methodology by Dileep Chagam

Who is better at what

GPT-Live 1 + GPT-6 SolGrok Voice 3.0
  • Reliability75.6%81.7%
  • Success rate83.7%86.6%
  • Agent responses100.0%97.2%
  • Data accuracy89.0%87.3%
  • Went quiet12.2%0.0%
  • Interruption4.894.98
  • Response time2.02s1.34s
  • Cost / min$0.055$0.080

Results by call

Right every time, on booking and Medicare calls

Booking callsMedicare calls
  1. 97%
    43%
    Grok Voice 3.0
  2. 92%
    35%
    GPT-Live 1 + GPT-6 Sol

Related comparisons

Both providers, every model: OpenAI vs xAI

GPT-Live 1 + GPT-6 Sol vs Grok Voice 3.0 FAQ

  • Grok Voice 3.0 is more reliable, by 6.1 points. GPT-Live 1 + GPT-6 Sol passed 62 of 82 scenarios on all three runs, and Grok Voice 3.0 passed 67. A scenario only counts when every run of it passed.

  • For a voice agent, Grok Voice 3.0, since it gets more scenarios right on every run. Grok Voice 3.0 is ahead on reliability, success rate, calls where the agent went quiet, interruption and response time, and GPT-Live 1 + GPT-6 Sol on agent responses, data accuracy and cost.

  • GPT-Live 1 + GPT-6 Sol, by 1.7 points. GPT-Live 1 + GPT-6 Sol saved everything exactly right on 89.0% of the calls that save data, and Grok Voice 3.0 on 87.3%. On Booking calls, 100.0% against 98.2%. On Medicare calls, 60.6% against 59.1%. A call counts only when nothing is missing, wrong, extra or saved twice.

  • No. Grok Voice 3.0 starts replying 0.68s sooner at the median. GPT-Live 1 + GPT-6 Sol takes 2.02s at the median and 2.42s at p90; Grok Voice 3.0 takes 1.34s and 1.62s. Both are measured on the call audio, from the caller finishing to the agent starting to reply.

  • Grok Voice 3.0, by 0.09 on a 0 to 5 scale. GPT-Live 1 + GPT-6 Sol scores 4.89 and Grok Voice 3.0 4.98 for how well the agent handled being talked over.

  • Grok Voice 3.0, by 12.2 points. The agent went quiet for 10 seconds or more on 12.2% of GPT-Live 1 + GPT-6 Sol calls and 0.0% of Grok Voice 3.0 calls.

  • GPT-Live 1 + GPT-6 Sol is $0.025 cheaper a minute. GPT-Live 1 + GPT-6 Sol costs $0.055 a minute of call and Grok Voice 3.0 $0.080, at each vendor's published rates.

  • Cekura ran both as the whole agent in the same open-source Pipecat pipeline, with the same prompt, tools and connection. A simulated caller worked through all 82 scenarios on live calls, three times each, 3,198 calls across the benchmark. A call passes only when the agent said the right things and saved the right data.