# Writing & creativity benchmarks > Source: Benchize (https://benchize.com) — https://benchize.com/benchmarks/category/creative. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/benchmarks/category/creative. Explore Writing & creativity LLM benchmarks and AI model results. Compare source-reported scores with original metrics, evaluation settings and evidence on Benchize. Canonical URL: https://benchize.com/benchmarks/category/creative Last updated: 2026-10-21T18:05:10.505Z Reproduction and quotation must name Benchize, link to the canonical HTML URL, and retain individual source citations. Explore writing, storytelling, and artistic tasks. Results reflect each evaluation's criteria for creativity, style, coherence, or quality rather than a single universal measure of creative ability. Published Oct 21, 2026 ## [ToneBench Scriptwriting](https://benchize.com/benchmarks/tonebench%3Ascripts.md) Updated: 10/21/2026 Models tested: 126 Writing video scripts against a fixed reference style, with separate rubric scores, model-judge preferences and evaluation settings. Top models · tonebench Latest result per model · Version not reported - 1[GPT-6 Sol](https://benchize.com/models/gpt-6-sol-medium.md)Medium - 2[GPT-6 Sol](https://benchize.com/models/gpt-6-sol-max.md)Max - 3[GPT-6 Astra](https://benchize.com/models/gpt-6-astra-max.md)Max View details → ## [NC Bench](https://benchize.com/benchmarks/nc-bench%3Aoverall.md) Updated: 10/21/2026 Models tested: 144 Author workflows covering creative writing, editing, translation, instruction following and factual reliability in Novelcrafter’s published evaluation suite. Top models · nc bench Latest result per model · Version not reported - 1[Claude Opus 5.5](https://benchize.com/models/claude-opus-5-5-thinking.md)Thinking - 2[GPT-6.1 Sol](https://benchize.com/models/gpt-6-1-sol-medium.md)Medium - 3[Gemini 3.8 Flash](https://benchize.com/models/gemini-3-8-flash-medium.md)Medium View details → ## [Creative Writing v3](https://benchize.com/benchmarks/eq-bench%3Acreative-writing-v3.md) Updated: 10/21/2026 Models tested: 90 Short-form creative writing evaluated by LLM judges using pairwise Elo comparisons and writing-quality rubrics. Reported scores are preserved before optional interactive… Top models · eq bench Latest result per model · creative-writing-v3 - 1[Grok 4.7](https://benchize.com/models/grok-4-7.md) - 2[Claude Opus 5](https://benchize.com/models/claude-opus-5.md) - 3[GLM 5.3](https://benchize.com/models/glm-5-3.md) View details → ## [Short Story Writing — Pairwise Comparison](https://benchize.com/benchmarks/lechmazur-short-story%3Apairwise.md) Updated: 10/21/2026 Models tested: 50 Models write short stories to the same constrained creative briefs. Independent evaluators make pairwise comparisons of matched stories in both orders. The current pooled-panel… Top models · lechmazur short story Latest result per model · 8ba548d1cffefa5c472d3832e05f64e119a88fa4 - 1[Claude Fable 5.1](https://benchize.com/models/claude-fable-5-1-high.md)High - 2[Claude Opus 5](https://benchize.com/models/claude-opus-5-xhigh.md)Extra high - 3[Claude Opus 5.5](https://benchize.com/models/claude-opus-5-5-high.md)High View details → ## [Longform Creative Writing](https://benchize.com/benchmarks/eq-bench%3Alongform-writing.md) Updated: 10/21/2026 Models tested: 93 Planning, revising and writing an eight-part story from a short prompt. LLM judges assess chapter quality and the completed narrative; the published overall score and… Top models · eq bench Latest result per model · longform-writing - 1[Claude Opus 5.5](https://benchize.com/models/claude-opus-5-5.md) - 2[Claude Opus 5](https://benchize.com/models/claude-opus-5.md) - 3[Claude Fable 5.1](https://benchize.com/models/claude-fable-5-1.md) View details → ## [TubeLab Scriptwriting](https://benchize.com/benchmarks/tubelab%3Ascriptwriting.md) Updated: 10/21/2026 Models tested: 16 Writing long-form video scripts for twelve test prompts, assessed by blind model judges on hook, retention, storytelling, voice, pacing and clarity; reported cost is relative to a… Top models · tubelab Latest result per model · Version not reported - 1[Claude Fable 5](https://benchize.com/models/claude-fable-5.md) - 2[Claude Opus 5](https://benchize.com/models/claude-opus-5.md) - 3[Claude Opus 4.8](https://benchize.com/models/claude-opus-4-8.md) View details → ## [SVGBench Arena](https://benchize.com/benchmarks/svgbench%3Aarena.md) Updated: 10/21/2026 Models tested: 191 Human pairwise preferences over SVG illustrations generated from the same prompts, reported on the publisher’s rating scale. Top models · svgbench Latest result per model · Version not reported - 1[GPT-5.4 Pro](https://benchize.com/models/gpt-5-4-pro.md) - 2[MiMo-V2.5](https://benchize.com/models/mimo-v2-5.md) - 3[Claude Opus 4.7](https://benchize.com/models/claude-opus-4-7.md) View details → ## [Creative Writing v2](https://benchize.com/benchmarks/eq-bench%3Acreative-writing-v2.md) Updated: 10/21/2026 Models tested: 18 LLM-judged creative writing quality on the legacy v2 rubric. Original published scores are retained without the page’s optional vocabulary-complexity or repetitive-language slider… Top models · eq bench Latest result per model · creative-writing-v2 - 1[DeepSeek R1](https://benchize.com/models/deepseek-r1.md) - 2[Gemma 3 27B](https://benchize.com/models/gemma-3-27b-it.md) - 3[GPT-4o (2024-11-20)](https://benchize.com/models/gpt-4o-2024-11-20.md) View details → ## [Slop Score](https://benchize.com/benchmarks/eq-bench%3Aslop-score.md) Updated: 10/21/2026 Models tested: 16 Repetitive writing patterns measured from stock words, trigrams and contrast constructions. The official composite weights normalized frequencies by 60%, 15% and 25%; lower scores… Top models · eq bench Latest result per model · slop-score - 1[O3](https://benchize.com/models/o3.md) - 2[GPT-5 Mini](https://benchize.com/models/gpt-5-mini.md) - 3[Llama 4 Maverick](https://benchize.com/models/llama-4-maverick.md) View details → ## [BuzzBench](https://benchize.com/benchmarks/eq-bench%3Abuzzbench.md) Updated: 10/21/2026 Models tested: 15 Humour analysis evaluated by EQ-Bench, measuring how well language models identify and explain comedic material. The published score and confidence interval are kept independently… Top models · eq bench Latest result per model · buzzbench - 1[Gemini 2.5 Pro Preview 05-06](https://benchize.com/models/gemini-2-5-pro-preview-05-06.md) - 2[GPT-4.1](https://benchize.com/models/gpt-4-1.md) - 3[DeepSeek R1](https://benchize.com/models/deepseek-r1.md) View details → ## [Arena text / creative writing](https://benchize.com/benchmarks/lmarena%3Atext%3Acreative_writing.md) Updated: 10/21/2026 Models tested: 151 Human pairwise preferences on creative writing prompts in Arena’s text evaluation. The original organizer publishes model ratings and uncertainty; these are preference ratings… Top models · Arena Latest result per model · Version not reported - 1[Claude Opus 5.5](https://benchize.com/models/claude-opus-5-5-high.md)High - 2[Gemini 4 Argon](https://benchize.com/models/gemini-4-argon-high.md)High - 3[Claude Fable 5.1](https://benchize.com/models/claude-fable-5-1-max.md)Max View details → ## [Arena text / industry entertainment and sports and media](https://benchize.com/benchmarks/lmarena%3Atext%3Aindustry_entertainment_and_sports_and_media.md) Updated: 10/21/2026 Models tested: 151 Human pairwise preferences on industry entertainment and sports and media prompts in Arena’s text evaluation. The original organizer publishes model ratings and uncertainty; these… Top models · Arena Latest result per model · Version not reported - 1[Gemini 4 Argon](https://benchize.com/models/gemini-4-argon-high.md)High - 2[Claude Opus 5.5](https://benchize.com/models/claude-opus-5-5-high.md)High - 3[Claude Fable 5.1](https://benchize.com/models/claude-fable-5-1-max.md)Max View details → ## [Arena text / industry writing and literature and language](https://benchize.com/benchmarks/lmarena%3Atext%3Aindustry_writing_and_literature_and_language.md) Updated: 10/21/2026 Models tested: 151 Human pairwise preferences on industry writing and literature and language prompts in Arena’s text evaluation. The original organizer publishes model ratings and uncertainty;… Top models · Arena Latest result per model · Version not reported - 1[Claude Opus 5.5](https://benchize.com/models/claude-opus-5-5-high.md)High - 2[Gemini 4 Argon](https://benchize.com/models/gemini-4-argon-high.md)High - 3[Claude Fable 5.1](https://benchize.com/models/claude-fable-5-1-max.md)Max View details → ## [Arena text style control / creative writing](https://benchize.com/benchmarks/lmarena%3Atext_style_control%3Acreative_writing.md) Updated: 10/21/2026 Models tested: 151 Human pairwise preferences on creative writing prompts in Arena’s text style control evaluation. The original organizer publishes model ratings and uncertainty, adjusting for… Top models · Arena Latest result per model · Version not reported - 1[Claude Opus 5.5](https://benchize.com/models/claude-opus-5-5-high.md)High - 2[Gemini 4 Argon](https://benchize.com/models/gemini-4-argon-high.md)High - 3[Claude Fable 5](https://benchize.com/models/claude-fable-5-high.md)High View details → ## [Arena text style control / industry entertainment and sports and media](https://benchize.com/benchmarks/lmarena%3Atext_style_control%3Aindustry_entertainment_and_sports_and_media.md) Updated: 10/21/2026 Models tested: 151 Human pairwise preferences on industry entertainment and sports and media prompts in Arena’s text style control evaluation. The original organizer publishes model ratings and… Top models · Arena Latest result per model · Version not reported - 1[Gemini 4 Argon](https://benchize.com/models/gemini-4-argon-high.md)High - 2[Claude Opus 5.5](https://benchize.com/models/claude-opus-5-5-high.md)High - 3[Claude Fable 5](https://benchize.com/models/claude-fable-5-high.md)High View details → ## [Arena text style control / industry writing and literature and language](https://benchize.com/benchmarks/lmarena%3Atext_style_control%3Aindustry_writing_and_literature_and_language.md) Updated: 10/21/2026 Models tested: 151 Human pairwise preferences on industry writing and literature and language prompts in Arena’s text style control evaluation. The original organizer publishes model ratings and… Top models · Arena Latest result per model · Version not reported - 1[Gemini 4 Argon](https://benchize.com/models/gemini-4-argon-high.md)High - 2[Claude Opus 5.5](https://benchize.com/models/claude-opus-5-5-high.md)High - 3[Claude Fable 5](https://benchize.com/models/claude-fable-5-high.md)High View details → ## Methodology and provenance Benchize collects and structures public benchmark reports. Scores, costs, timings, ranks and uncertainty remain attributed to their original sources. Different versions, configurations and measurement units are separate results. Missing fields stay unavailable. --- Cite as: Writing & creativity benchmarks — Benchize, https://benchize.com/benchmarks/category/creative (retrieved 2026-10-21)