Creative writing
Writing stories and longer fiction. In the tests here an AI judge either compares models' stories head to head or scores them against a list of what good writing does.
Sources last checked 2026-10-23.
We can't show scores for this kind of work yet. EQ-Bench (Sam Paech) publishes results from two tests. See who tests it
At a glance
Vendor claims have not been collected for any category yet.
ⓘ More about this category
- Reliability measured by a linked source; figures not shown here
- 2 sources, linked results only
- Sources last checked 2026-10-28
- Ring 4: needs rings 1 to 3 first
- EQCW = EQ-Bench Creative Writing v3, EQLW = EQ-Bench Longform Creative Writing
Tested by
EQ-Bench Creative Writing v3
EQ-Bench (Sam Paech), checked 2026-10-28Each model writes short stories for 32 set prompts, and an AI judge compares them head to head with other models' stories to rank writing quality.
Last updated: not published by the tester. We last saw current models listed on 2026-10-28.Covers 30 of the models we trackScores not shown: no license for reuse stated
See their resultsMore about this test: EQ-Bench Creative Writing v3
Reliability: EQ-Bench runs each of the 32 writing prompts three times per model (96 items).
Technical details for EQ-Bench Creative Writing v3
Elo score: Each model writes short stories for 32 set prompts, and an AI judge compares them head to head with other models' stories to rank writing quality.
License: Score reuse not stated. Checked 2026-10-28.
EQ-Bench Longform Creative Writing
EQ-Bench (Sam Paech), checked 2026-10-28Each model plans and writes a story of about eight chapters, and an AI judge scores each chapter and the whole piece against a list of what good writing does.
Last updated: not published by the tester. We last saw current models listed on 2026-10-28.Covers 31 of the models we trackScores not shown: no license for reuse stated
See their resultsMore about this test: EQ-Bench Longform Creative Writing
Reliability not measured by this source.
Technical details for EQ-Bench Longform Creative Writing
Score (0-100): Each model plans and writes a story of about eight chapters, and an AI judge scores each chapter and the whole piece against a list of what good writing does.
License: Score reuse not stated. Checked 2026-10-28.
What this doesn't tell you
- Quality is judged by an AI judge's preferences, not by a right answer.
- These tests cover fiction only, not business writing.
- Scores from different sources are not comparable, even when they share a unit.
- None of these tests were run on your own documents, code or data.
- No source for this task allows us to show its figures yet. Follow the links for their results.
Data version 2026-10-28+91ac3fc9d071 · Terms of use