Gemini 3 Flash

Google · Gemini · measured on live video

Overshoot API id gemini-3-flash-preview

Real-time under 200 msClosedImage + Video

Speed, TTFT, E2E, and price at Google Vertex, US East (Virginia) reference region. Rank is position on the VLM Intelligence Index.

Capability breakdown

Fastest loop: Overshoot · US East (Virginia) · 139 ms end to end.


Overshoot	US East (Virginia)	254 tok/s	45 ms	139 ms	$4.66	<200 ms
Overshoot	US West (Oregon)	254 tok/s	51 ms	145 ms	$4.66	<200 ms
Google Vertex	US East (Virginia)	247 tok/s	53 ms	150 ms	$4.52	<200 ms
Overshoot	EU Central (Frankfurt)	254 tok/s	63 ms	158 ms	$4.66	<200 ms
Overshoot	EU West (Ireland)	254 tok/s	64 ms	159 ms	$4.66	<200 ms
Google Vertex	US West (Oregon)	247 tok/s	69 ms	166 ms	$4.52	<200 ms
Google Vertex	EU West (Ireland)	247 tok/s	76 ms	173 ms	$4.52	<200 ms
Overshoot	Asia Pacific (Tokyo)	254 tok/s	82 ms	176 ms	$4.66	<200 ms
Overshoot	Asia Pacific (Mumbai)	254 tok/s	98 ms	192 ms	$4.66	<200 ms
Overshoot	Middle East (Dubai)	254 tok/s	98 ms	193 ms	$4.66	<200 ms
Google Vertex	Asia Pacific (Tokyo)	247 tok/s	96 ms	194 ms	$4.52	<200 ms
Overshoot	South America (São Paulo)	254 tok/s	105 ms	199 ms	$4.66	<200 ms
Google Vertex	Asia Pacific (Mumbai)	247 tok/s	106 ms	204 ms	$4.52	–

Illustrative samples shared across models, not verbatim output from Gemini 3 Flash.

Scene description

“Describe what's happening in this camera frame.”

A forklift is reversing toward loading bay 3 while a worker in a hi-vis vest signals from the left. Two pallets remain unwrapped near the dock edge.

OCR + extraction

“Read the shipping label and return structured JSON.”

{ "tracking": "1Z A24 9W7 03 8421 990 4", "carrier": "UPS", "weight_kg": 12.4, "dest": "Portland, OR 97204" }

Grounding

“Point to every person not wearing a hard hat.”

2 detections, [x:0.41, y:0.62, w:0.08, h:0.19] conf 0.94 · [x:0.77, y:0.55, w:0.07, h:0.21] conf 0.88.

Gemini 3.1 Provs

Gemma 4 31BReal-timevs

Gemma 4 26B A4BReal-timevs