BlogLLM Benchmark Scores as Published, Side by Side
PinchBench: Results as Published
Every result the catalog holds for PinchBench: 14 models from 8 developers, read from 3 pages. Each number is as printed by the company named beside it; nothing is ranked or rescored. Models are listed by developer, newest first.
Area Tool Use and WorkflowsMaintainer Not statedIntroduced 2026Data version 2026103001, as of 2026-10-30
PinchBench is printed on the NVIDIA Nemotron 3 Ultra model card among the agentic benchmarks; the card does not describe it and lists it among the evaluations run on an official implementation or an internal scaffolding. As remembered, it scores an assistant agent on everyday tool-use tasks run through the Pinch harness.
Not checked against a maintainer's page.
PinchBench
Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.