Back to Gap Map Model components / Benchmark / eval datasets

IFEval

open / Overall score: 4.0(strong)

Instruction-Following Eval: 541 prompts carrying about 25 types of verifiable instruction - word-count limits, keyword inclusion, format requirements - so that compliance can be scored programmatically rather than by a judge model. It is a core component of the Hugging Face Open LLM Leaderboard.

Openness

5 high confidence

5.0

license: Apache-2.0(OSI/open,redistributable)
access: public(not gated)
datasheet: present(card+structure+citation)
splits: public(single train split, no held-out)

Fully open: Apache-2.0, ungated, redistributable, with a complete dataset card.

https://huggingface.co/datasets/google/IFEval recorded 2026-09-29
Dataset card tagged "license:apache-2.0" with repo state "gated":false, whose summary reads "This dataset contains the prompts used in the Instruction-Following Eval (IFEval) benchmark" and whose sections include "Data Splits" and "Citation Information".

Adoption

4 high confidence

4.0

Adoption is measured by Hugging Face downloads of google/IFEval. IFEval is a core component of the Hugging Face Open LLM Leaderboard and the standard benchmark for instruction-following evaluation.

https://huggingface.co/api/datasets/google/IFEval recorded 2026-08-12
116,029 downloads in the trailing 30 days for google/IFEval

Capability

not assessed

A dataset is not 'capable', so capability is left unscored; openness and adoption carry this category.

https://huggingface.co/datasets/google/IFEval recorded 2026-09-29
The card summary reads "This dataset contains the prompts used in the Instruction-Following Eval (IFEval) benchmark for large language models." Searched the body for performance, throughput, accuracy and SOTA claims: none.

Verified 2026-09-27