← Claude Sonnet 5 · API · High effort (default)
Evidenceexcludedindependent_lab
IFBench
Artificial Analysis model card — IFBench field null for Claude Sonnet 5 max
ev_2026_07_sonnet5_ifbench_null
Result
0
Confidence D
- Tested
- 2026-07-11
Configuration
How it was measured
- Exact model
- claude-sonnet-5
- Reasoning
- max
- Agent harness
- none
- Tool access
- n/a
- Prompting
- n/a — metric not published
Explicit null record: Artificial Analysis lists ifbench=null for Claude Sonnet 5 (max). Prevents fabricated IFBench scores. Re-check AA later.
Caveats
Known limitations
- This is a negative finding (missing metric), not a zero capability claim.
Weighting
Evidence weight factors
The engine weights each evidence row by these factors (0–1) when fusing scores — higher is more trusted.
relevance0.50
quality1.00
recency1.00
health0.90
comparability0.50
reliability0.30
Benchmark
IFBench
Harder verifiable instruction-following constraints designed to test generalization beyond saturated IFEval-style benchmarks.
Benchmark source →