CekuraBench
All comparisons
OpenAICascade baseline

GPT Realtime 2.1 vs Flux → GPT-4.1 → ElevenLabs Flash

Flux → GPT-4.1 → ElevenLabs Flash is ahead on reliability, success rate, data accuracy and stalled calls, and GPT Realtime 2.1 on interruption and response time. The two tie on agent responses.

Last updated Methodology by Dileep Chagam

Compare

Success rate by suite

Higher is better
Appointments59 scenariosMedicare23 scenarios
  • GPT Realtime 2.1OpenAI
    92%
    81%
  • Cascade, GPT-4.1Not ranked
    97%
    83%
Each model's share of calls that passed every check, on each suite.
Verdict

Which to pick

Pick GPT Realtime 2.1 for handling interruptions (0.02 higher) and fast replies (0.23s sooner). Pick Flux → GPT-4.1 → ElevenLabs Flash for reliability across repeated calls (3.7 points higher), calls that pass every check (4.1 points higher), saving the caller's data correctly (2.1 points higher) and fewer stalled calls (2.4 points lower). They are level on getting the agent's replies right.

Every metric, side by side

Reliability

Scenarios that passed on all three runs.

GPT Realtime 2.1
79.3%1st
Cascade, GPT-4.1
82.9%

Cascade, GPT-4.1, 3.7 points higher

Success rate

Calls that passed every check.

GPT Realtime 2.1
89.0%1st
Cascade, GPT-4.1
93.1%

Cascade, GPT-4.1, 4.1 points higher

Agent responses

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

GPT Realtime 2.1
99.2%2nd
Cascade, GPT-4.1
99.2%

Level

Data accuracy

Calls where everything the agent saved matched exactly.

GPT Realtime 2.1
92.0%1st
Cascade, GPT-4.1
94.1%

Cascade, GPT-4.1, 2.1 points higher

Stalled calls

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

GPT Realtime 2.1
3.3%5th
Cascade, GPT-4.1
0.8%

Cascade, GPT-4.1, 2.4 points lower

Interruption

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

GPT Realtime 2.1
4.984th
Cascade, GPT-4.1
4.96

GPT Realtime 2.1, 0.02 higher

Response time

Median time to start replying after the caller finishes.

GPT Realtime 2.1
1.94s5th
Cascade, GPT-4.1
2.17s

GPT Realtime 2.1, 0.23s sooner

Cost / min

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

GPT Realtime 2.1
$0.0856th
Cascade, GPT-4.1
No verified rate

Not comparable

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

GPT Realtime 2.1 answers first

GPT Realtime 2.1 1.94s, Cascade, GPT-4.1 2.17s at the median.

Response time, median to p90

Lower is betterMedianTo p90
Medianp90
  • GPT Realtime 2.1OpenAI
    GPT Realtime 2.1: median 1.94s, p90 2.46s
    1.94s2.46s
  • Cascade, GPT-4.1Not ranked
    Cascade, GPT-4.1: median 2.17s, p90 2.51s
    2.17s2.51s
2.0s3.0s
Seconds from the caller finishing to the agent starting to reply, per model.

What we saw on the calls

GPT Realtime 2.1

  • Almost never fails, and rarely twice in a row.
All 4 notes on GPT Realtime 2.1
Model
gpt-realtime-2.1
Setting
Reasoning high
Voice
marin
Turn taking
Service turn detection
Audio
24 kHz
Cost
Excludes caller transcription, which the service does not meter per call.

Flux → GPT-4.1 → ElevenLabs Flash

  • Most scenarios pass every run, and none fails every run.
All 4 notes on Cascade, GPT-4.1
Model
flux-general-en → gpt-4.1 → eleven_flash_v2_5
Setting
No reasoning setting
Voice
ElevenLabs 21m00Tcm4TlvDq8ikWAM
Turn taking
Speech-to-text turn detection
Audio
16 kHz
Cost
Not published: no verified combined rate for the three services.
See the full leaderboardReference agent

Related comparisons

Frequently asked questions

Is GPT Realtime 2.1 or Flux → GPT-4.1 → ElevenLabs Flash more reliable?

Flux → GPT-4.1 → ElevenLabs Flash is more reliable, by 3.7 points. GPT Realtime 2.1 passed 65 of 82 scenarios on all three runs, and Flux → GPT-4.1 → ElevenLabs Flash passed 68. A scenario only counts when every run of it passed.

Is GPT Realtime 2.1 better than a cascaded STT, LLM and TTS pipeline?

On reliability, no: the Flux → GPT-4.1 → ElevenLabs Flash cascade passes more scenarios on every run. Flux → GPT-4.1 → ElevenLabs Flash is ahead on reliability, success rate, data accuracy and stalled calls, and GPT Realtime 2.1 on interruption and response time. The two tie on agent responses. Both ran as the same agent, with the same prompt and tools.

Which saves caller data more accurately, GPT Realtime 2.1 or Flux → GPT-4.1 → ElevenLabs Flash?

Flux → GPT-4.1 → ElevenLabs Flash, by 2.1 points. GPT Realtime 2.1 saved everything exactly right on 92.0% of the calls that save data, and Flux → GPT-4.1 → ElevenLabs Flash on 94.1%. On Appointments, 95.9% against 98.8%. On Medicare, 81.8% against 81.8%. A call counts only when nothing is missing, wrong, extra or saved twice.

Is GPT Realtime 2.1 faster than Flux → GPT-4.1 → ElevenLabs Flash?

Yes. GPT Realtime 2.1 starts replying 0.23s sooner at the median. GPT Realtime 2.1 takes 1.94s at the median and 2.46s at p90; Flux → GPT-4.1 → ElevenLabs Flash takes 2.17s and 2.51s. Both are measured on the call audio, from the caller finishing to the agent starting to reply.

Which handles interruptions better, GPT Realtime 2.1 or Flux → GPT-4.1 → ElevenLabs Flash?

GPT Realtime 2.1, by 0.02 on a 0 to 5 scale. GPT Realtime 2.1 scores 4.98 and Flux → GPT-4.1 → ElevenLabs Flash 4.96 for how well the agent handled being talked over.

Which goes silent on callers less often, GPT Realtime 2.1 or Flux → GPT-4.1 → ElevenLabs Flash?

Flux → GPT-4.1 → ElevenLabs Flash, by 2.4 points. The agent went silent for 10 seconds or more on 3.3% of GPT Realtime 2.1 calls and 0.8% of Flux → GPT-4.1 → ElevenLabs Flash calls.

Which is cheaper, GPT Realtime 2.1 or Flux → GPT-4.1 → ElevenLabs Flash?

There is no verified per-minute rate for Flux → GPT-4.1 → ElevenLabs Flash, so cost cannot be compared yet.

Where does the GPT Realtime 2.1 vs Flux → GPT-4.1 → ElevenLabs Flash data come from?

Cekura ran both as the whole agent in the same open-source Pipecat pipeline, with the same prompt, tools and connection. A simulated caller worked through all 82 scenarios on live calls, three times each, 2,952 calls across the benchmark. A call passes only when the agent said the right things and saved the right data. The cascade runs the same agent for reference and is never ranked.