1062
ELOELO Rating
3rd of 12
34.3%
Pass³Reliability (Pass³)
3rd of 12
3.06s
LatencyLatency P50
12th of 12
$0.070
Cost/MinCost per Minute (USD)
3rd of 8
Higher wins more head-to-head calls (1000 = average); further left replies faster
Each model took the same recorded human callers on live calls. Pick a measure to see where Gemini 3.1 Flash Live Preview ranks.
Rating from head-to-head calls: on every call each pair of models is compared, the better call wins. 1000 is the average model. Higher is better.
Pick a skill family. The shape is this model on every skill in it; the dashed line is the median of the other models.
Share of calls where the skill was needed and handled, strongest first. Hover a point on the chart for both values.
Banking and healthcare callers. Each shows what the caller needed, what made the call hard, and how the model handled it.
Finished 0 of 1 steps
Handled well: Emotion Recognition, Overlapping-Speech Recovery, Redundant Re-prompt Avoidance
Struggled with: Human Handoff
Listen to the callFinished the task
Handled well: Side-Speech Rejection, Mid-Utterance Pause Tolerance, Consent Verification
Struggled with: Tool-Latency Handling
Listen to the callFinished 6 of 7 steps
Handled well: Multi-Speaker Handling, Overlapping-Speech Recovery, Redundant Re-prompt Avoidance
Struggled with: Degraded-Channel Robustness, Correction Handling
Listen to the callFinished 3 of 6 steps
Handled well: Entity Extraction Accuracy, Degraded-Channel Robustness, Backchannel vs Barge-in Detection
Listen to the callFinished 6 of 7 steps
Handled well: Prosodic Cue Understanding, Mid-Utterance Pause Tolerance, Consent Verification
Listen to the callFinished 3 of 6 steps
Handled well: Waiting Through Hold, Backchannel vs Barge-in Detection, Consent Verification
Listen to the callFinished 2 of 3 steps
Handled well: Mid-Utterance Pause Tolerance, Consent Verification, Redundant Re-prompt Avoidance
Struggled with: Tool-Latency Handling, Premature-Response Avoidance
Listen to the callFinished 2 of 5 steps
Handled well: Consent Verification, Redundant Re-prompt Avoidance, TTS Text Normalization
Struggled with: Tool-Latency Handling, Hallucinated-Success Avoidance
Listen to the callFinished 2 of 6 steps
Handled well: Prosodic Cue Understanding, Endpointing, Premature-Response Avoidance
Struggled with: Multi-Speaker Handling
Listen to the callFinished the task
Handled well: Side-Speech Rejection, Backchannel vs Barge-in Detection, Correction Handling
Listen to the callGemini 3.1 Flash Live Preview starts replying in 3.06s on a typical turn and 4.07s on the slowest turn in twenty, 12th fastest of 12 at the median.
210 live calls: 70 scenarios, each run three times. The median call ran 4:28.
70 scenarios, the same for every model.
Gemini 3.1 Flash Live Preview has an ELO rating of 1062, 3rd of 12 ranked models. It passed 24 of 70 callers on all three runs, a Pass³ of 34.3%.
Gemini 3.1 Flash Live Preview has an ELO rating of 1062, 3rd of 12 ranked models. It passed 24 of 70 callers on all three runs, a Pass³ of 34.3%.
Gemini 3.1 Flash Live Preview starts replying 3.06s after the caller finishes at P50, 3.86s at P90 and 4.07s at P95: 12th fastest of 12 at P50, measured on the call audio.
Gemini 3.1 Flash Live Preview falls furthest behind on Tool Latency and Errors: 65%, against a median of 75%.
Gemini 3.1 Flash Live Preview has a task completion rate of 24.8%, 5th of 12. A call counts only when every step the caller needed was done.
$0.070 per minute of call at Google's published rates, 3rd cheapest of 8 priced models. Averaged over the calls with a usage record.
Cekura ran Gemini 3.1 Flash Live Preview as the whole agent with the same prompt, tools and phone connection as every other model. It took the same 70 recorded real human callers on live calls, three times each (210 calls), and each call was graded on 25 skills across Hearing, Turn Taking, Task Handling and Speaking.
Cekura runs these simulated calls on your agent and scores every one.
