Deepslate Opal vs Gemini 3.8 Live

Gemini 3.8 Live is more reliable, by 14.3 points. Deepslate Opal handled 9 of 70 real callers well on all three runs, and Gemini 3.8 Live 19. A caller only counts when every run of the call scored 80 or more.

For a voice agent, Gemini 3.8 Live, since it handles more real callers well on every run. Gemini 3.8 Live is ahead on ELO rating, turn taking score, speaking score, task handling score, reliability, calls handled well, calls fully resolved, calls where the agent went quiet and median latency, and Deepslate Opal on hearing score.

Gemini 3.8 Live, by 3.8 points higher. Deepslate Opal fully resolved 12.4% of calls and Gemini 3.8 Live 16.2%. A call counts only when every step the caller needed was done.

No. Gemini 3.8 Live starts replying 0.45s sooner at the median. Deepslate Opal takes 2.46s at the median and 3.16s at p90; Gemini 3.8 Live takes 2.01s and 2.52s. Both are measured on the call audio, from the caller finishing to the agent starting to reply.

Gemini 3.8 Live, by 10.0 points. The agent went quiet for 10 seconds or more on 16.2% of Deepslate Opal calls and 6.2% of Gemini 3.8 Live 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.