GPT Realtime 2.1 vs Phonic-1

GPT Realtime 2.1 is more reliable, by 1.4 points. GPT Realtime 2.1 handled 23 of 70 real callers well on all three runs, and Phonic-1 22. 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, speaking score, reliability, calls handled well and calls fully resolved, and Phonic-1 on turn taking score, task handling score, calls where the agent went quiet and median latency.

GPT Realtime 2.1, by 9.5 points higher. GPT Realtime 2.1 fully resolved 34.3% of calls and Phonic-1 24.8%. A call counts only when every step the caller needed was done.

No. Phonic-1 starts replying 0.17s sooner at the median. GPT Realtime 2.1 takes 2.16s at the median and 2.70s at p90; Phonic-1 takes 1.99s and 2.32s. Both are measured on the call audio, from the caller finishing to the agent starting to reply.

Phonic-1, by 1.0 points. The agent went quiet for 10 seconds or more on 11.9% of GPT Realtime 2.1 calls and 11.0% of Phonic-1 calls.

There is no verified per-minute rate for Phonic-1, so cost cannot be compared yet.

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.