Grok Voice Think Fast 2.0 vs GPT Realtime 2.1

GPT Realtime 2.1 is more reliable, by 4.3 points. Grok Voice Think Fast 2.0 handled 20 of 70 real callers well on all three runs, and GPT Realtime 2.1 23. A caller only counts when every run of the call scored 80 or more.

For a voice agent, GPT Realtime 2.1, since it handles more real callers well on every run. GPT Realtime 2.1 is ahead on ELO rating, hearing score, task handling score, reliability, calls handled well and calls fully resolved, and Grok Voice Think Fast 2.0 on turn taking score, speaking score, calls where the agent went quiet, median latency and cost.

GPT Realtime 2.1, by 11.4 points higher. Grok Voice Think Fast 2.0 fully resolved 22.9% of calls and GPT Realtime 2.1 34.3%. A call counts only when every step the caller needed was done.

Yes. Grok Voice Think Fast 2.0 starts replying 0.72s sooner at the median. Grok Voice Think Fast 2.0 takes 1.44s at the median and 1.83s at p90; GPT Realtime 2.1 takes 2.16s and 2.70s. Both are measured on the call audio, from the caller finishing to the agent starting to reply.

Grok Voice Think Fast 2.0, by 3.3 points. The agent went quiet for 10 seconds or more on 8.6% of Grok Voice Think Fast 2.0 calls and 11.9% of GPT Realtime 2.1 calls.

Grok Voice Think Fast 2.0 is $0.005 cheaper a minute. Grok Voice Think Fast 2.0 costs $0.080 a minute of call and GPT Realtime 2.1 $0.085, at each vendor's published rates.

Cekura ran both as the whole agent with the same prompt, tools and phone connection. Both took the same 70 recorded real human callers on live calls, three times each. Each call is graded on 25 skills across Hearing, Turn Taking, Task Handling and Speaking, and passes at a score of 80 out of 100.