VECTOR WIREAI INTELLIGENCE
⌘K
PKT
Refresh Models —Deals —Regulatory —Sources —

Questions this page answers

Which model leads IFBench (prompt)?
Across 5 models scored on IFBench (prompt), Granite 4.2 8B leads at 79.3, ahead of Granite 4.2 30B at 74.7. The median tracked score is 74.3, and the field spans 71.9 to 79.3.

IFBench (prompt)

Across 5 models scored on IFBench (prompt), Granite 4.2 8B leads at 79.3, ahead of Granite 4.2 30B at 74.7. The median tracked score is 74.3, and the field spans 71.9 to 79.3.

5 models tracked
Data as of August 25, 2026
#ModelVendorBest scoreRunsLast seen
1Granite 4.2 8BIBM79.322026-08-25
2Granite 4.2 30BIBM74.712026-08-25
3Granite 4.2 3BIBM74.322026-08-25
4Nemotron 3 Super 120B A12BNVIDIA73.412026-07-07
5NVIDIA-Nemotron-Labs-3-Puzzle-75B-A9B-BF16NVIDIA71.912026-07-07

Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.