Claude Sonnet 5 vs Claude Sonnet 5.5

Best cell per row highlighted. Null means undisclosed — never counted as zero.

Anthropic
Claude Sonnet 5GA
8.9
AI Panel
Anthropic
Claude Sonnet 5.5GA
9.3
AI Panel

Identity & lifecycle

ProviderAnthropicAnthropic
Family / tierClaude 5 · SonnetClaude 5 · Sonnet
StatusGAGA
Released2026-06-302026-09-28
Knowledge cutoff2026-012026-06

Architecture & context

Context window1M1M
Max output tokens128K128K
Input modalitiestext, imagetext, image
Reasoning modeoptionaloptional
Open weightsNoNo

Pricing (per Mtok)

Input$2$2
Output$10$10
Cached input$0.2$0.2
Batch input$1$1
Free tierYesYes

Speed

Output speed (tok/s)7197.6
Time to first token (s)11.712.7
Latency tierfastfast

Trust & deployment

Trains on inputsNoNo
LicenseProprietaryProprietary
CloudsBedrock, Vertex AI, Azure AI FoundryBedrock, Vertex AI, Azure AI Foundry
ComplianceSOC2, HIPAA, ISO27001, ISO42001SOC2, HIPAA, ISO27001, ISO42001

AI Panel scoring

Unified score8.99.3
Decision Maker99.5
Domain Strategist8.59.2
Finance Lead88.8
Domain Practitioner9.59.3
Power User99.2
Skeptic7.58
Value score8.59.2

Benchmarks

GPQA Diamond91.1—
SWE-bench Verified85.2—
Terminal-Bench80.4—
Humanity's Last Exam43.256.9
LMArena Elo14611471
LMArena Coding Elo15201536
Artificial Analysis Index3856

Scores link to their sources. Missing cells mean the vendor hasn't disclosed a result — honesty over padding.