UCP PlaygroundGrok 3 Mini (Reasoning)/Models/Grok 3 Mini (Reasoning)
InspectorAgentDocs
xAIxAI

Grok 3 Mini (Reasoning) Retired

This model is no longer available to run. Its stats below reflect past sessions and are kept for reference.

xAI's compact reasoning model. Inherited the Grok line's reliability (near-zero failures) but with lower checkout completion (25%), tending to stop at search results. Retired: deprecated upstream in favour of the Grok 4 line.

Avg Tokens35,632Avg Duration40.6sTurns to Checkout8.6
CheckoutCartSearchTurnsTokens11.113.348.94322
Shopping Score22/100
Weak

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%11.1%
Cart Rate20%13.3%
Search Rate10%48.9%
Turn Efficiency15%43%
Token Efficiency15%22%
51.1% of sessions failed with errors.

Top Stores

StoreCheckout %Cart %
katko.com50%50%
allbirds.com25%25%
houseofparfum.nl25%25%
everlane.com0%0%
mikesbikes.com0%0%
••••••••••••••••••••••••••••••••0%0%
kyliecosmetics.com0%0%

Known Issues

No known issues documented for this model yet.

Token Usage

Avg per Session35,632
Fleet Average66,685
vs Fleet−47%
Median (p50)13,164
p90116,513
Prompt / Completion split
Prompt 33,173 (93%) Completion 2,458 (7%)
No daily trend data yet. Run more sessions to see token usage over time.
Distribution range
Min0p250p5013,164p7534,970p90116,513Max248,956

Test Grok 3 Mini (Reasoning) on Your Store

Run a live agent session to see how Grok 3 Mini (Reasoning) handles your store's checkout flow end-to-end.

Run Agent Session

Compare With Other Models

MetaMetaLlama 3.3 70B39.3%GoogleGoogleGemini 3.5 Flash38.9%AnthropicAnthropicClaude Sonnet 4.636.7%xAIxAIGrok 4.326.3%GoogleGoogleGemini 2.5 Flash25.9%AnthropicAnthropicClaude Opus 4.823.2%DeepSeekDeepSeekDeepSeek V3.222.1%GoogleGoogleGemini 3.1 Pro21.1%GoogleGoogleGemini 2.5 Pro20%OpenAIOpenAIGPT-4o15.1%OpenAIOpenAIGPT-5.511.1%xAIxAIGrok 3 Mini (Reasoning)11.1%OpenAIOpenAIo4-mini (Reasoning)9.5%AnthropicAnthropicClaude Sonnet 59.4%DeepSeekDeepSeekDeepSeek V4 Pro8.8%DeepSeekDeepSeekDeepSeek V4 Flash6.5%Moonshot AIMoonshot AIKimi K34.5%xAIxAIGrok 4.54.2%DeepSeekDeepSeekDeepSeek R1 (Reasoning)4.1%GoogleGoogleGemini 3.6 Flash3.8%MetaMetaLlama 4 Maverick2.7%AnthropicAnthropicClaude Opus 51.8%AlibabaAlibabaQwen 3.7 Max0%AlibabaAlibabaQwQ 32B (Reasoning)0%
Full comparison view →