CekuraBench
All comparisons
GoogleGoogle

Gemini 3.1 Flash Live Preview vs Gemini 3.8 Live Extended Thinking

Gemini 3.1 Flash Live Preview is ahead on reliability, agent responses, stalled calls and interruption, and Gemini 3.8 Live Extended Thinking on success rate, data accuracy and response time.

Compare

Success rate by suite

Appointments59 scenariosMedicare23 scenarios
  • GPT Realtime 2.1OpenAI
    92%
    81%
  • GPT-Live 1 + Sol lowOpenAI
    97%
    49%
  • Gemini 3.8 Live ThinkingGoogle
    82%
    61%
  • Grok Think Fast 2.0xAI
    96%
    25%
  • Gemini 3.1 Flash LiveGoogle
    93%
    32%
  • Phonic v1Phonic
    97%
    14%
  • Realtime 2.1 MiniOpenAI
    84%
    4%
  • Nova 2 SonicAmazon
    80%
    6%
  • Cascade, GPT-4.1Not ranked
    97%
    83%
Each model's share of calls that passed every check, on each suite.

Every metric, side by side

Reliability

Scenarios that passed on all three runs.

Gemini 3.1 Flash Live
62.2%5th
Gemini 3.8 Live Thinking
56.1%6th

Gemini 3.1 Flash Live, 6.1 points higher

Success rate

Calls that passed every check.

Gemini 3.1 Flash Live
76.0%4th
Gemini 3.8 Live Thinking
76.4%3rd

Gemini 3.8 Live Thinking, 0.4 points higher

Agent responses

Calls where the agent's replies reached the expected outcome.

Gemini 3.1 Flash Live
95.9%5th
Gemini 3.8 Live Thinking
94.7%6th

Gemini 3.1 Flash Live, 1.2 points higher

Data accuracy

Calls where everything the agent saved matched exactly.

Gemini 3.1 Flash Live
79.2%4th
Gemini 3.8 Live Thinking
80.2%3rd

Gemini 3.8 Live Thinking, 0.9 points higher

Stalled calls

Calls where the agent went silent for 10 seconds or more.

Gemini 3.1 Flash Live
3.3%5th
Gemini 3.8 Live Thinking
3.7%6th

Gemini 3.1 Flash Live, 0.4 points lower

Interruption

How well it handled being talked over, 0 to 5.

Gemini 3.1 Flash Live
4.984th
Gemini 3.8 Live Thinking
4.906th

Gemini 3.1 Flash Live, 0.08 higher

Response time

Median time to start replying after the caller finishes.

Gemini 3.1 Flash Live
2.88s8th
Gemini 3.8 Live Thinking
2.32s7th

Gemini 3.8 Live Thinking, 0.56s sooner

Cost / min

One minute of call at the vendor's published rates.

Gemini 3.1 Flash Live
$0.0703rd
Gemini 3.8 Live Thinking
No verified rate

Not comparable

The small figure is the place among ranked models. The cascade is not ranked.

Gemini 3.8 Live Thinking answers first

Gemini 3.8 Live Thinking 2.32s, Gemini 3.1 Flash Live 2.88s at the median.

Response time, median to p90

MedianTo p90
Medianp90
  • Phonic v1Phonic
    Phonic v1: median 1.59s, p90 1.86s
    1.59s1.86s
  • Grok Think Fast 2.0xAI
    Grok Think Fast 2.0: median 1.62s, p90 1.94s
    1.62s1.94s
  • GPT Realtime 2.1OpenAI
    GPT Realtime 2.1: median 1.94s, p90 2.46s
    1.94s2.46s
  • GPT-Live 1 + Sol lowOpenAI
    GPT-Live 1 + Sol low: median 2.02s, p90 2.42s
    2.02s2.42s
  • Realtime 2.1 MiniOpenAI
    Realtime 2.1 Mini: median 2.08s, p90 2.75s
    2.08s2.75s
  • Nova 2 SonicAmazon
    Nova 2 Sonic: median 2.18s, p90 2.76s
    2.18s2.76s
  • Gemini 3.8 Live ThinkingGoogle
    Gemini 3.8 Live Thinking: median 2.32s, p90 3.06s
    2.32s3.06s
  • Gemini 3.1 Flash LiveGoogle
    Gemini 3.1 Flash Live: median 2.88s, p90 4.00s
    2.88s4.00s
  • Cascade, GPT-4.1Not ranked
    Cascade, GPT-4.1: median 2.17s, p90 2.51s
    2.17s2.51s
2.0s3.0s4.0s
Seconds from the caller finishing to the agent starting to reply, per model.

What we saw on the calls

Gemini 3.1 Flash Live Preview

  • Steady on the short form, though it repeats lookups it already made.
  • On the long form it hands the caller over with the wrong details.
  • Slowest to reply; after several tool calls its replies fall behind or go missing.
  • Sometimes says goodbye without hanging up.
Model
models/gemini-3.1-flash-live-preview
Setting
Thinking level high
Voice
Charon
Turn taking
Shared local turn detection
Audio
16 kHz
Cost
Averaged over the calls with a usage record.

Gemini 3.8 Live Extended Thinking

  • Strong on the long form; when it slips, it skips a save it said it made.
  • Weak on the short form: phone numbers saved wrong, lookups never made.
  • Acts on half an answer instead of asking; weakest when the caller corrects themselves.
  • Least predictable: the same scenario passes one run and fails the next.
  • Replies fall behind or stop; callers hang up waiting.
Model
gemini-3.8-live-extended-thinking
Setting
Thinking level high
Voice
Charon
Turn taking
Shared local turn detection
Audio
16 kHz
Cost
Some calls used cached input the vendor meters at a rate we have not verified.
See the full leaderboardReference agent

Related comparisons

Frequently asked questions

Is Gemini 3.1 Flash Live Preview or Gemini 3.8 Live Extended Thinking more reliable?

Gemini 3.1 Flash Live Preview passed 51 of 82 scenarios on all three runs, and Gemini 3.8 Live Extended Thinking passed 46. Gemini 3.1 Flash Live Preview is ahead, by 6.1 points.

Which saves caller data more accurately, Gemini 3.1 Flash Live Preview or Gemini 3.8 Live Extended Thinking?

Gemini 3.1 Flash Live Preview saved everything exactly right on 79.2% of the calls that save data, and Gemini 3.8 Live Extended Thinking on 80.2%. On Appointments, 96.5% against 86.5%. On Medicare, 34.8% against 63.6%. A call counts only when nothing is missing, wrong, extra or saved twice.

Which responds faster, Gemini 3.1 Flash Live Preview or Gemini 3.8 Live Extended Thinking?

Gemini 3.8 Live Extended Thinking starts replying 0.56s sooner at the median. Gemini 3.1 Flash Live Preview takes 2.88s at the median and 4.00s at p90; Gemini 3.8 Live Extended Thinking takes 2.32s and 3.06s. Both are measured on the call audio, from the caller finishing to the agent starting to reply.

Which costs less, Gemini 3.1 Flash Live Preview or Gemini 3.8 Live Extended Thinking?

There is no verified per-minute rate for Gemini 3.8 Live Extended Thinking, so cost cannot be compared yet.

How were Gemini 3.1 Flash Live Preview and Gemini 3.8 Live Extended Thinking compared?

Both ran as the whole agent in the same open-source Pipecat pipeline, with the same prompt, tools and phone line. A simulated caller worked through all 82 scenarios on live calls, three times each, 2,214 calls across the benchmark. A call passes only when the agent said the right things and saved the right data.