UCP PlaygroundGemini 3.5 Flash/Models/Gemini 3.5 Flash
InspectorAgentDocs
GoogleGoogle

Gemini 3.5 Flash

Google's fast tier (Gemini 3.5 Flash, succeeding the 3 Flash preview). The 3 Flash generation was the cleanest tool caller in our benchmark — just 1.0% error rate — fast and with a strong 58% checkout rate. The 3.5 revision is newer and being re-benchmarked. The pick when you want reliable results without frontier cost.

Avg Tokens78,945Avg Duration23.2sTurns to Checkout7.9
CheckoutCartSearchTurnsTokens385680.1511
Shopping Score42/100
Fair

Scores computed from real agent sessions against live UCP-enabled stores. Not estimated — every data point is from an actual tool call, checkout attempt, and store response.

Shopping Score Breakdown

Checkout Rate40%38%
Cart Rate20%56%
Search Rate10%80.1%
Turn Efficiency15%51%
Token Efficiency15%1%
19.6% of sessions failed with errors.

Top Stores

StoreCheckout %Cart %
forever21.com100%100%
houseofparfum.nl68.4%82.4%
monos.com66.7%66.7%
oakywood.shop66.7%66.7%
ucp.travel53.8%54.4%
••••••••••••••••50%66.7%
••••••••••••••••50%75%
sini.fi50%50%
••••••••••••••••50%100%
allbirds.com21.4%35.7%

Known Issues

Turn Countlow

8.8 avg turns to checkout — could be more efficient despite clean tool calling.

Token Usage

Avg per Session78,945
Fleet Average74,995
vs Fleet+5%
Median (p50)38,556
p90157,057
Prompt / Completion split
Prompt 77,100 (98%) Completion 1,314 (2%)
Daily Average Tokens (last 30 days)
Distribution range
Min0p2520,200p5038,556p7574,577p90157,057Max1,392,682

Test Gemini 3.5 Flash on Your Store

Run a live agent session to see how Gemini 3.5 Flash handles your store's checkout flow end-to-end.

Run Agent Session

Compare With Other Models

MetaMetaLlama 3.3 70B39.3%GoogleGoogleGemini 3.5 Flash38%AnthropicAnthropicClaude Sonnet 4.634.7%xAIxAIGrok 4.326.3%GoogleGoogleGemini 2.5 Flash23.3%AnthropicAnthropicClaude Opus 4.822.3%DeepSeekDeepSeekDeepSeek V3.220.7%AlibabaAlibabaQwen 3.8 Flash20%GoogleGoogleGemini 2.5 Pro19.8%GoogleGoogleGemini 3.1 Pro19.5%OpenAIOpenAIGPT-4o15%DeepSeekDeepSeekDeepSeek V4.1 Flash14.3%GoogleGoogleGemini 3.8 Flash12.5%xAIxAIGrok 3 Mini (Reasoning)11.1%OpenAIOpenAIGPT-5.510.2%OpenAIOpenAIo4-mini (Reasoning)9.1%AnthropicAnthropicClaude Opus 58.9%xAIxAIGrok 4.58.1%AnthropicAnthropicClaude Sonnet 57.7%DeepSeekDeepSeekDeepSeek V4 Pro7.6%DeepSeekDeepSeekDeepSeek V4 Flash7.5%DeepSeekDeepSeekDeepSeek R1 (Reasoning)3.5%Moonshot AIMoonshot AIKimi K33%AlibabaAlibabaQwen 3.7 Max2.6%MetaMetaLlama 4 Maverick2.2%GoogleGoogleGemini 3.6 Flash1.1%GoogleGoogleGemini 3.7 Flash0%OpenAIOpenAIGPT-6 Luna0%MetaMetaMuse Spark 1.30%AlibabaAlibabaQwQ 32B (Reasoning)0%
Full comparison view →