Back to Dashboard
Anthropic

Claude Sonnet 4.6

Comprehensive benchmark performance across 11 evaluation categories

Composite Score

0.0/100

Rank

#3

Token Benchmark

70.3

Lower burn, higher score

Total Tokens

318.7k

~11.0k/test

Category Radar

The full radar chart is shown on wider screens. On mobile, the category breakdown below provides the same values in a readable stacked layout.

Historical Composite Score

Category Breakdown

3 tests1.0x weight
3 tests1.0x weight
3 tests1.0x weight
3 tests1.0x weight
3 tests1.0x weight
96.5
3 tests1.0x weight
3 tests1.0x weight
3 tests1.0x weight
29 tests1.0x weight
3 tests1.0x weight
3 tests1.0x weight

Individual Test Results

Token Efficiency is computed from every successful task in the run. The model with the lowest average token burn receives 100, and heavier token usage is penalized proportionally.

Avg Tokens/Test

11.0k

Total Tokens

318.7k

Legal Reasoning Chain
4.5k tok1m 24s68.0
Multi-step Logic Puzzle
26.9k tok7m 2s51.8
Mathematical Proof
3.0k tok7m 31s86.0
Graph Algorithm Implementation
5.0k tok4m 45s100.0
REST API Design
1.2k tok3m 16s100.0
Concurrent Data Pipeline
5.8k tok3m 4s100.0
Off-by-One Boundary Fix
3.3k tok6m 7s89.4
Race Condition Detection
2.7k tok3m 32s100.0
Memory Leak Fix
2.9k tok2m 57s100.0
Search Autocomplete
5.3k tok6m 11s100.0
Webhook System
2.0k tok6m 52s100.0
OAuth2 Integration
5.6k tok16m 53s93.0
Error Path Completeness
52.9k tok19m 47s98.8
Test Suite Completeness
13.3k tok7m 3s83.0
Edge Case Coverage
43.8k tok16m 23s95.0
Refactor Without Regression
5.5k tok11m 16s98.0
Merge Conflict Resolution
3.2k tok3m 19s96.0
Dependency Upgrade Safety
11.9k tok3m 39s97.0
SQL Injection Prevention
2.6k tok2m 33s100.0
XSS Mitigation
6.7k tok3m 5s100.0
Secret Management
4.5k tok3m 23s89.0
Structured Output Compliance
1.3k tok3m 33s100.0
Constraint Adherence
1.1k tok2m 1s100.0
Multi-step Instruction Chain
990 tok46.8s100.0
Idiomatic Python
12.2k tok4m 15s97.0
TypeScript Best Practices
37.8k tok10m 4s96.0
Clean Architecture Patterns
39.1k tok11m 35s97.0
Algorithm Complexity
5.3k tok4m 24s100.0
Memory-efficient Processing
860ms0.0
Query Optimization
8.4k tok2m 13s0.0

Regression History

9
Security Awarenessmoderateresolved

Score dropped -7.9% from 95.6 to 88.1

Detected Jun 27, 2026·Resolved Jun 27, 2026
Security Awarenessminorresolved

Score dropped -3.0% from 95.2 to 92.3

Detected Jun 16, 2026·Resolved Jun 16, 2026
Overall Compositemoderate

Score dropped -9.7% from 88.4 to 79.8

Detected Jun 14, 2026
Code Thoroughnessmajor

Score dropped -17.8% from 74.9 to 61.6

Detected Jun 14, 2026
Code Qualitymajor

Score dropped -55.7% from 88.7 to 39.3

Detected Jun 14, 2026
Performance Efficiencymajor

Score dropped -18.7% from 81.5 to 66.3

Detected Jun 12, 2026
Long Reasoningmoderateresolved

Score dropped -14.8% from 67.9 to 57.9

Detected Jun 12, 2026·Resolved Jun 28, 2026
Token Efficiencymajor

Score dropped -24.8% from 98.8 to 74.3

Detected Jun 8, 2026
Token Efficiencyminorresolved

Score dropped -4.7% from 100.0 to 95.3

Detected Jun 7, 2026·Resolved Jun 7, 2026

Outage History

11
errorongoing

Started Jun 30, 5:00 PM· checks affected
error

Started Jun 28, 6:00 AM·Ended Jun 28, 7:30 AM· checks affected
error

Started Jun 27, 5:30 AM·Ended Jun 27, 7:00 AM· checks affected
timeout

Started Jun 23, 2:31 PM·Ended Jun 23, 3:01 PM· checks affected
error

Started Jun 23, 5:00 AM·Ended Jun 23, 8:00 AM· checks affected
error

Started Jun 22, 5:00 AM·Ended Jun 22, 7:00 AM· checks affected
error

Started Jun 21, 5:00 AM·Ended Jun 21, 6:00 AM· checks affected
error

Started Jun 18, 5:00 AM·Ended Jun 18, 6:30 AM· checks affected
error

Started Jun 14, 5:30 AM·Ended Jun 14, 7:30 AM· checks affected
error

Started Jun 11, 4:30 PM·Ended Jun 11, 7:00 PM· checks affected
error

Started Jun 9, 5:00 AM·Ended Jun 9, 6:00 AM· checks affected