← leaderboard
Claude Fable 5
SUSPICIOUSAnthropic · anthropic/claude-fable-5 · verdict as of 2026-08-10
Flags
- · instruction-following: score 90.0 vs baseline 100.0 (z=-5.00, drop 10.0%)
Daily test history (90 days, baseline band = trailing mean ± 2σ)
Math & Reasoning
NOT COOKED
Structured Extraction
NOT COOKED
Game Design
no data
Code Generation
NOT COOKED
Instruction Following
SUSPICIOUS
Report Analysis
NOT COOKED
Summarization
NOT COOKED
Customer Service
NOT COOKED
Web Design
NOT COOKED
Creative Writing
no data
Serving providers (last 30 days)
striped = multiple providers served this model that day (hover for detail)
Public benchmarks overall 97.0
MMLU-Pro
—
GPQA Diamond
92.6
SWE-bench Verified
—
LMArena Elo
1508
AIME 2025
—
retrieved 2026-07-03 from public sources — see methodology
Recent samples (latest run, one per test case)
No samples from the latest run.