Grok Voice 3.0vsGPT Realtime 2.1 Mini

Grok Voice 3.0 is more reliable, by 31.7 points. Grok Voice 3.0 passed 67 of 82 scenarios on all three runs, and GPT Realtime 2.1 Mini passed 41. 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, agent responses, data accuracy, calls where the agent went quiet and response time, and GPT Realtime 2.1 Mini on interruption and cost.

Grok Voice 3.0, by 22.8 points. Grok Voice 3.0 saved everything exactly right on 87.3% of the calls that save data, and GPT Realtime 2.1 Mini on 64.5%. On Booking calls, 98.2% against 88.8%. On Medicare calls, 59.1% against 0.0%. A call counts only when nothing is missing, wrong, extra or saved twice.

Yes. Grok Voice 3.0 starts replying 0.74s sooner at the median. Grok Voice 3.0 takes 1.34s at the median and 1.62s at p90; GPT Realtime 2.1 Mini takes 2.08s and 2.75s. Both are measured on the call audio, from the caller finishing to the agent starting to reply.

GPT Realtime 2.1 Mini, by 0.01 on a 0 to 5 scale. Grok Voice 3.0 scores 4.98 and GPT Realtime 2.1 Mini 4.99 for how well the agent handled being talked over.

Grok Voice 3.0, by 2.5 points. The agent went quiet for 10 seconds or more on 0.0% of Grok Voice 3.0 calls and 2.5% of GPT Realtime 2.1 Mini calls.

GPT Realtime 2.1 Mini is $0.060 cheaper a minute. Grok Voice 3.0 costs $0.080 a minute of call and GPT Realtime 2.1 Mini $0.020, 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.