Deepslate Opal vs GPT Realtime 2.1 Mini

GPT Realtime 2.1 Mini is more reliable, by 1.4 points. Deepslate Opal handled 9 of 70 real callers well on all three runs, and GPT Realtime 2.1 Mini 10. A caller only counts when every run of the call scored 80 or more.

For a voice agent, GPT Realtime 2.1 Mini, since it handles more real callers well on every run. GPT Realtime 2.1 Mini is ahead on ELO rating, turn taking score, task handling score, reliability, calls handled well, calls fully resolved, calls where the agent went quiet, median latency and cost, and Deepslate Opal on hearing score. The two tie on speaking score.

GPT Realtime 2.1 Mini, by 1.4 points higher. Deepslate Opal fully resolved 12.4% of calls and GPT Realtime 2.1 Mini 13.8%. A call counts only when every step the caller needed was done.

No. GPT Realtime 2.1 Mini starts replying 0.06s sooner at the median. Deepslate Opal takes 2.46s at the median and 3.16s at p90; GPT Realtime 2.1 Mini takes 2.40s and 2.88s. Both are measured on the call audio, from the caller finishing to the agent starting to reply.

GPT Realtime 2.1 Mini, by 7.1 points. The agent went quiet for 10 seconds or more on 16.2% of Deepslate Opal calls and 9.0% of GPT Realtime 2.1 Mini calls.

GPT Realtime 2.1 Mini is $0.092 cheaper a minute. Deepslate Opal costs $0.112 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 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.