UCP PlaygroundClaude Sonnet 4.6/Models/Claude Sonnet 4.6
InspectorAgentDocs
AnthropicAnthropic

Claude Sonnet 4.6

Anthropic's prior-generation balanced model (Sonnet 4.6, superseded by Sonnet 5), kept as a lower-cost option. Strong multi-turn checkout flow with wallet payment and identity linking support; higher token usage than Opus due to verbose tool responses.

Avg Tokens70,823Avg Duration34.1sTurns to Checkout8.4
CheckoutCartSearchTurnsTokens36.948.379.8450
Shopping Score39/100
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%36.9%
Cart Rate20%48.3%
Search Rate10%79.8%
Turn Efficiency15%45%
Token Efficiency15%0%
19.5% of sessions failed with errors.

Top Stores

StoreCheckout %Cart %
••••••••••••••••100%100%
houseofparfum.nl75.5%85.9%
forever21.com66.7%66.7%
•••••••••••••••••••62.5%62.5%
••••••••••••••••••••••••••••••••54.5%54.5%
••••••••••••••••••••••••••••••••50%100%
mr-fothergills.co.uk50%50%
westwing.de50%50%
reebok.com50%50%
••••••••••••••••••••••••••••••••50%100%

Known Issues

Nudge Dependencymedium

Stops after validation errors 9.3% of the time. Needs orchestrator nudge to continue checkout.

High Token Usagelow

Averages 67K tokens per session — highest among Anthropic models. Verbose tool response handling.

Meta Stringificationmedium

Frequently stringifies meta objects instead of passing as JSON. Requires orchestrator auto-injection for idempotency keys.

Token Usage

Avg per Session70,823
Fleet Average66,593
vs Fleet+6%
Median (p50)39,866
p90133,329
Prompt / Completion split
Prompt 69,459 (98%) Completion 1,065 (2%)
Daily Average Tokens (last 30 days)
Distribution range
Min0p2519,558p5039,866p7577,502p90133,329Max1,935,784

Test Claude Sonnet 4.6 on Your Store

Run a live agent session to see how Claude Sonnet 4.6 handles your store's checkout flow end-to-end.

Run Agent Session

Compare With Other Models

MetaMetaLlama 3.3 70B39.3%GoogleGoogleGemini 3.5 Flash39%AnthropicAnthropicClaude Sonnet 4.636.9%GoogleGoogleGemini 2.5 Flash26.5%xAIxAIGrok 4.326.3%AnthropicAnthropicClaude Opus 4.823.3%DeepSeekDeepSeekDeepSeek V3.222.1%GoogleGoogleGemini 3.1 Pro21.3%GoogleGoogleGemini 2.5 Pro20.1%OpenAIOpenAIGPT-4o15.6%OpenAIOpenAIGPT-5.511.3%xAIxAIGrok 3 Mini (Reasoning)11.1%AnthropicAnthropicClaude Sonnet 511.1%OpenAIOpenAIo4-mini (Reasoning)9.6%DeepSeekDeepSeekDeepSeek V4 Pro9%Moonshot AIMoonshot AIKimi K35%xAIxAIGrok 4.54.5%DeepSeekDeepSeekDeepSeek R1 (Reasoning)4.1%DeepSeekDeepSeekDeepSeek V4 Flash3.4%MetaMetaLlama 4 Maverick2.8%AnthropicAnthropicClaude Opus 52%GoogleGoogleGemini 3.6 Flash2%AlibabaAlibabaQwen 3.7 Max0%AlibabaAlibabaQwQ 32B (Reasoning)0%
Full comparison view →