CekuraBench

Gemini 3.8 Live Extended Thinking

Every modelCompare

52.4%

Reliability

9th of 12

2.07s

Median reply

8th of 12

Reliability vs. response time

Higher and further left is better

Best quadrant
Reliability, pass³ (%)
Median response time (s)

For the wider field, see our pick of the best voice AI APIs for real-time audio.

Gemini 3.8 Live Extended Thinking against every other model

Scenarios that passed on all 3 runs, pass³. Higher is better.

#ModelReliability
  1. 01Grok Voice 3.081.7%
  2. 02GPT Realtime 2.179.3%
  3. 03GPT-Live 1 + GPT-6 Sol75.6%
  4. 04Azure Realtime70.7%
  5. 05Phonic v168.3%
  6. 06Grok Voice Think Fast 2.067.1%
  7. 07Gemini 3.1 Flash Live Preview62.2%
  8. 08Deepslate Opal57.3%
  9. 09Gemini 3.8 Live Extended Thinking52.4%
  10. 10Amazon Nova 2 Sonic50.0%
  11. 11GPT Realtime 2.1 Mini50.0%
  12. 12Gemini 3.8 Live41.5%
  • refCascade82.9%

Where Gemini 3.8 Live Extended Thinking struggled

Gemini 3.8 Live Extended Thinking falls furthest behind on accent and slow speech: 50% passed, against a median of 83%.

Below the medianPassedMedian
  • Accent & slow speech
    50%median 83%
  • Plain booking & lookup
    76%median 94%
  • Tool or system errors
    75%median 92%
  • Caller corrects or changes mind
    67%median 81%
  • Background noise
    67%median 81%
  • Silence, cough & packet loss
    83%median 88%

Response time at P50, P90 and P95

Gemini 3.8 Live Extended Thinking starts replying in 2.07s on a typical turn and 2.93s on the slowest turn in twenty, 8th fastest of 12 at the median.

ModelP50P90P95
  • Cascade
    2.17s2.51s2.62s
Seconds from the caller finishing to the agent replying. The bar is the typical turn, P50.

Gemini 3.8 Live Extended Thinking head to head

Every comparisonAll 12 models

How much audio Gemini 3.8 Live Extended Thinking was tested on

246 live calls: 82 scenarios, each run three times.

Live calls
246
Scenarios
82
Booking calls
177
Medicare calls
69
Replies per call
9.0

82 scenarios, the same for every model. See every test case

Gemini 3.8 Live Extended Thinking FAQ

  • Gemini 3.8 Live Extended Thinking passed 43 of 82 scenarios on all three runs, 52.4%, 9th of 12 realtime models. Across all 246 calls, 76.0% passed every check.

  • Gemini 3.8 Live Extended Thinking starts replying 2.07s after the caller finishes at the median, 2.74s at P90 and 2.93s at P95. That is 8th fastest of 12 at the median, measured on the call audio.

  • Gemini 3.8 Live Extended Thinking falls furthest behind the field on accent and slow speech: 50% of calls passed, against a median of 83%.

  • Gemini 3.8 Live Extended Thinking saved everything exactly right on 84.4% of the calls that save data, 5th of 12. On Booking calls, 88.9%. On Medicare calls, 72.7%.

  • There is no verified per-minute rate for Gemini 3.8 Live Extended Thinking, so it is left out of the cost ranking.

  • Cekura ran Gemini 3.8 Live Extended Thinking as the whole agent in the same open-source Pipecat pipeline as every other model, with the same prompt, tools and connection. A simulated caller worked through 82 scenarios on live calls, three times each: 246 calls for Gemini 3.8 Live Extended Thinking. A call passes only when the agent said the right things and saved the right data.