Gemini 3.8 Live vs GPT Realtime 2.1

GPT Realtime 2.1 is more reliable, by 5.7 points. Gemini 3.8 Live handled 19 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, turn taking score, task handling score, reliability, calls handled well and calls fully resolved, and Gemini 3.8 Live on speaking score, calls where the agent went quiet and median latency.

GPT Realtime 2.1, by 18.1 points higher. Gemini 3.8 Live fully resolved 16.2% of calls and GPT Realtime 2.1 34.3%. A call counts only when every step the caller needed was done.

Yes. Gemini 3.8 Live starts replying 0.15s sooner at the median. Gemini 3.8 Live takes 2.01s at the median and 2.52s 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.

Gemini 3.8 Live, by 5.7 points. The agent went quiet for 10 seconds or more on 6.2% of Gemini 3.8 Live calls and 11.9% of GPT Realtime 2.1 calls.

There is no verified per-minute rate for Gemini 3.8 Live, 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.