model radar platform
Claude Code Radar
Track Claude Code model score, cost, runtime, cache behavior, and quota posture in a format that stays tied to onesagent product decisions.
platform signal
measured snapshotClaude Code
Anthropic coding runtime tracked via public Claude Code Radar data
Claude Code public benchmark snapshot
Current public data emphasizes single-snapshot benchmark, cost, time, and cache metrics rather than event windows.
Model-level score and cache posture
Cross-model decision support
Source freshness and attribution
Direct handoff into onesagent model choice
Current snapshot
Generated: 2026-07-01T17:55:13+08:00
Latest successful sync: 2026-07-01T17:55:13+08:00
Freshness: fresh_with_public_source_sync
Quota radar
Quota summary
Quota capacity by visible tier or shared bucket.
Observed usage posture
unknown
5h
unknown
7d
unknown
Live model table
The models buyers are actually comparing.
One list, current metrics, no extra model subpages.
Model
Score
Cost
Runtime
Tasks passed
Cache hit
Effort
Trend
Sonnet 5 Max
yellowScore:120
Cost:$125.10
Runtime:8.4h
Tasks passed:8/10
Cache hit:unknown
Effort:unknown
Trend:-
Sonnet 5 High
yellowScore:62.5
Cost:$54.70
Runtime:3.4h
Tasks passed:5/12
Cache hit:unknown
Effort:unknown
Trend:-
Cross-platform check
Open Codex Radar
Need the other provider? Open its live table instead of drilling into stale subpages.
Latest status
Current public data emphasizes single-snapshot benchmark, cost, time, and cache metrics rather than event windows.