Best LLMs for Writing in 2026

Aggregated benchmark data across EQ-Bench Creative Writing, LMArena Text, and Artificial Analysis — covering 52 models, updated weekly.

Last updated: October 12, 2026 · 52 models tracked · 3 tiers: Premium · Mid-Range · Budget

The short answer

Updated October 2026

The best LLM for writing right now is GPT-6 Astra (OpenAI), which tops the EQ-Bench Creative Writing leaderboard ahead of Claude Fable 5.1 and Claude Opus 5. For the best quality per dollar, GLM-5.3 (Zhipu AI) delivers near-frontier writing at a fraction of the cost. Full ranking across 52 models below.

1 GPT-6 Astra OpenAI EQ Creative 2173
2 Claude Fable 5.1 Anthropic EQ Creative 2162
3 Claude Opus 5 Anthropic EQ Creative 2133

How we rank models for writing

This ranking combines three independent data sources to give the most complete picture of writing quality across frontier LLMs. No single benchmark captures the full picture — so we aggregate:

Show benchmark details

EQ Creative EQ-Bench Creative Writing — specialist benchmark using trained raters to assess narrative quality, emotional depth, prose style, and character voice. Elo scale ~1300–2175. The most relevant signal for marketing copy, long-form content, and creative work.
Arena Text LMArena Text — crowd-sourced human preference leaderboard. Broad signal across all text tasks: a model that consistently wins votes is generally pleasant, clear, and useful to read. Elo scale ~1346–1507.
EQ General EQ-Bench General — measures emotional intelligence in roleplay scenarios. A proxy for character voice quality and tonal control — useful for brand voice work. Note: high EQ-General does not automatically mean strong creative writing; interpret alongside EQ Creative.
Speed Artificial Analysis — median output tokens per second across providers. Matters for iterative draft workflows where waiting costs time. ~75 tokens ≈ 55 words.

Prices are per 1M tokens (input / output) and reflect standard API pricing. A dash (—) means the model has not yet appeared on that leaderboard — never an estimated or interpolated value.

Model	EQ Creative ↓	Arena Text ↕	EQ General ↕	Speed ↕	Price / 1M ↕
GPT-6 Astra Premium OpenAI 📋 Consensus	2,173.3	1,475	—	—	$10.00 $50.00
Sources EQ-Bench leaderboard → LMArena Text →
Claude Fable 5.1 Premium Anthropic 📋 Consensus	2,162	1,501	—	—	$10.00 $50.00
Sources EQ-Bench leaderboard → LMArena Text →
Claude Opus 5 Premium Anthropic 📋 Consensus	2,132.6	1,490	—	—	$5.00 $25.00
Sources EQ-Bench leaderboard → LMArena Text →
GPT-6 Sol Premium OpenAI 📋 Consensus	2,124.7	1,456	—	—	$2.00 $10.00
Sources EQ-Bench leaderboard → LMArena Text →
Kimi K3 Premium Moonshot AI 📋 Consensus	2,082.3	1,488	—	—	$3.00 $15.00
Sources EQ-Bench leaderboard → LMArena Text →
GLM-5.3 Mid-Range Zhipu AI 📋 Consensus	2,075	1,478	—	—	$1.40 $4.40
Sources EQ-Bench leaderboard → LMArena Text →
Claude Opus 5.5 Premium Anthropic 📋 Consensus	2,050.1	1,507	—	—	$4.00 $20.00
Sources EQ-Bench leaderboard → LMArena Text →
GPT-5.6 Sol Premium OpenAI 📋 Consensus	1,971.6	1,485	—	—	$5.00 $30.00
Sources EQ-Bench leaderboard → LMArena Text →
Claude Fable 5 Premium Anthropic 📋 Consensus	1,943.1	1,504	2,049.7	—	$10.00 $50.00
Sources EQ-Bench leaderboard → LMArena Text →
Muse Spark 1.1 Mid-Range Meta 📋 Consensus	1,927.1	1,491	—	—	$1.25 $4.25
Sources EQ-Bench leaderboard → LMArena Text →
Claude Opus 4.7 Premium Anthropic 📋 Consensus	1,914.3	1,501	1,884.3	—	$5.00 $25.00
Sources EQ-Bench leaderboard → LMArena Text →
Muse Spark 1.3 Mid-Range Meta 📋 Consensus	1,905.9	1,494	—	—	$1.25 $4.25
Sources EQ-Bench leaderboard → LMArena Text →
GPT-5.6 Terra Premium OpenAI 📋 Consensus	1,854.9	1,466	—	—	$2.00 $12.00
Sources EQ-Bench leaderboard → LMArena Text →
GPT-5.5 Premium OpenAI 📋 Consensus	1,843.9	1,482	1,576.9	—	$5.00 $30.00
Sources EQ-Bench leaderboard → LMArena Text →
Muse Spark 1.2 Mid-Range Meta 📋 Consensus	1,840.4	1,492	—	—	$1.25 $4.25
Sources EQ-Bench leaderboard → LMArena Text →
Claude Opus 4.8 Premium Anthropic 📋 Consensus	1,839.8	1,481	2,029.8	—	$5.00 $25.00
Sources EQ-Bench leaderboard → LMArena Text →
GPT-5.4 Premium OpenAI 📋 Consensus	1,839.7	1,475	1,562.9	—	—
Sources EQ-Bench leaderboard → LMArena Text →
GPT-5.6 Luna Budget OpenAI 📋 Consensus	1,828.7	1,452	—	—	$0.20 $1.20
Sources EQ-Bench leaderboard → LMArena Text →
Claude Sonnet 4.6 Premium Anthropic 📋 Consensus	1,810.4	1,472	1,714.1	50 t/s	$3.00 $15.00
Sources EQ-Bench leaderboard → LMArena Text → Artificial Analysis →
Claude Opus 4.6 Premium Anthropic 📋 Consensus	1,809.1	1,504	1,717.4	45 t/s	$5.00 $25.00
Sources EQ-Bench leaderboard → LMArena Text → Artificial Analysis →
Claude Sonnet 5 Premium Anthropic 📋 Consensus	1,794	1,461	—	—	$3.00 $15.00
Sources EQ-Bench leaderboard → LMArena Text →
GLM-5.2 Mid-Range Zhipu AI 📋 Consensus	1,756.5	1,475	1,575.1	—	$1.40 $4.40
Sources EQ-Bench leaderboard → LMArena Text →
Gemini 3.8 Flash Mid-Range Google 📋 Consensus	1,747.9	1,497	—	—	$0.75 $3.75
Sources EQ-Bench leaderboard → LMArena Text →
Kimi K2.6 Mid-Range Moonshot AI 📋 Consensus	1,724.5	1,461	1,561.2	—	$0.60 $2.50
Sources EQ-Bench leaderboard → LMArena Text →
Gemini 3.7 Flash Mid-Range Google 📋 Consensus	1,723.2	1,486	—	—	$1.50 $7.50
Sources EQ-Bench leaderboard → LMArena Text →
GPT-5.2 Premium OpenAI 📋 Consensus	1,702.9	1,437	1,558.8	—	$1.25 $10.00
Sources EQ-Bench leaderboard → LMArena Text →
GPT-5.3 Chat Mid-Range OpenAI 🧠 EQ-Bench	1,689.9	—	1,393	—	—
Sources EQ-Bench leaderboard →
Claude Opus 4.5 Premium Anthropic 📋 Consensus	1,686.7	1,474	1,545.7	—	$5.00 $25.00
Sources EQ-Bench leaderboard → LMArena Text →
Claude Sonnet 4.5 Premium Anthropic 📋 Consensus	1,677.6	1,456	1,511	—	$3.00 $15.00
Sources EQ-Bench leaderboard → LMArena Text →
O3 Mid-Range OpenAI 📋 Consensus	1,675.6	1,432	1,500	—	$2.00 $8.00
Sources EQ-Bench leaderboard → LMArena Text →
Kimi K2 Mid-Range Moonshot AI 🧠 EQ-Bench	1,665.7	—	1,562.2	44 t/s	$0.55 $2.20
Sources EQ-Bench leaderboard → Artificial Analysis →
Horizon Alpha Mid-Range Unknown 🧠 EQ-Bench	1,622.7	—	1,554.4	—	—
Sources EQ-Bench leaderboard →
Gemini 3.6 Flash Mid-Range Google 📋 Consensus	1,603.5	1,482	—	—	$1.50 $7.50
Sources EQ-Bench leaderboard → LMArena Text →
GLM-5 Budget Zhipu AI 📋 Consensus	1,600.6	1,458	1,526	80 t/s	$0.80 $2.50
Sources EQ-Bench leaderboard → LMArena Text → Artificial Analysis →
GLM-5.1 Mid-Range Zhipu AI 📋 Consensus	1,592.2	1,464	1,566.5	—	—
Sources EQ-Bench leaderboard → LMArena Text →
Claude Opus 4 Premium Anthropic 📋 Consensus	1,579.7	1,426	1,399.8	—	$15.00 $75.00
Sources EQ-Bench leaderboard → LMArena Text →
Kimi K2.5 Mid-Range Moonshot AI 📋 Consensus	1,578.6	1,450	1,544.7	—	—
Sources EQ-Bench leaderboard → LMArena Text →
DeepSeek V4 Pro Budget DeepSeek 📋 Consensus	1,553.2	1,465	1,570.2	—	$0.43 $0.87
Sources EQ-Bench leaderboard → LMArena Text →
Gemini 3 Pro Mid-Range Google 📋 Consensus	1,524.5	1,485	1,559.4	80 t/s	$2.00 $12.00
Sources EQ-Bench leaderboard → LMArena Text → Artificial Analysis →
DeepSeek V3.2 Budget DeepSeek 📋 Consensus	1,515	1,425	—	—	$0.28 $0.42
Sources EQ-Bench leaderboard → LMArena Text →
Gemini 3.1 Pro Mid-Range Google 📋 Consensus	1,491.3	1,487	1,537.7	—	$2.00 $12.00
Sources EQ-Bench leaderboard → LMArena Text →
Mistral Medium 3 Budget Mistral AI 🧠 EQ-Bench	1,475.7	—	—	—	$0.40 $2.00
Sources EQ-Bench leaderboard →
Qwen3-235B Budget Alibaba 📋 Consensus	1,459	1,375	1,213.6	—	$0.18 $0.54
Sources EQ-Bench leaderboard → LMArena Text →
GPT-4o Premium OpenAI 📋 Consensus	1,443	1,346	1,393	185 t/s	$2.50 $10.00
Sources EQ-Bench leaderboard → LMArena Text → Artificial Analysis →
Gemini 2.5 Pro Premium Google 📋 Consensus	1,421.2	1,445	1,545	—	$1.25 $10.00
Sources EQ-Bench leaderboard → LMArena Text →
GLM-4.7 Budget Zhipu AI 📋 Consensus	1,413.3	1,441	1,428.3	—	$0.38 $1.70
Sources EQ-Bench leaderboard → LMArena Text →
MiniMax M2.5 Budget MiniMax 📋 Consensus	1,361	1,391	—	395 t/s	$0.30 $1.20
Sources EQ-Bench leaderboard → LMArena Text → Artificial Analysis →
Gemini 3 Flash Mid-Range Google 🏟️ Arena	—	1,473	—	250 t/s	$0.50 $3.00
Sources LMArena Text → Artificial Analysis →
Gemini 3.1 Flash-Lite Budget Google 🏟️ Arena	—	1,433	—	—	$0.25 $1.50
Sources LMArena Text →
Grok 4.1 Mid-Range xAI 🏟️ Arena	—	1,465	—	163 t/s	$0.20 $0.50
Sources LMArena Text → Artificial Analysis →
MiMo-V2.5 Budget Xiaomi 🏟️ Arena	—	1,434	—	—	$0.40 $2.00
Sources LMArena Text →
GPT-5.1 Premium OpenAI 📋 Consensus	—	1,456	1,551	—	$1.25 $10.00
Sources EQ-Bench leaderboard → LMArena Text →

The right model depends on the task

Benchmark leaderboards rank models globally — but the best model for a 2,000-word thought leadership article is not necessarily the best model for a 15-word social media headline. Here's how the leading models split across common writing tasks:

Narrative & long-form

Thought leadership, case studies, email newsletters, ghostwriting. Requires emotional depth, tonal consistency, and the ability to sustain voice across thousands of words.

Best picks: GPT-6 Astra · Claude Sonnet 4.6

Structured commercial copy

Product descriptions, landing pages, ad copy, LinkedIn posts. Requires clarity, persuasion structure, and format adherence more than creative flair.

Best picks: GPT-5.5 · Claude Sonnet 4.6

High-volume / fast drafts

Social media scheduling, meta descriptions, bulk content variation. Speed and cost matter more than peak quality; fast iteration wins here.

Best picks: Gemini 3 Flash · Grok 4.1 · Kimi K2

Brand voice & consistency

Any content where staying on-brand is non-negotiable. Requires strong instruction-following, tonal control, and memory of brand guidelines.

Best picks: Claude Opus 4.8 · Claude Sonnet 4.6

Managing this by hand means juggling several API keys, pricing tiers, and a decision tree for every task type. The table above shows where each model wins, so you can match the model to the job instead of forcing one model onto everything. For analysis, finance and other business work, see the best AI for business work ranking.

The model sets the ceiling. The harness decides how much of it you use.

Two people on the same model get different writing. The model is the engine. What changes the outcome is what surrounds it: whether it already knows your voice and your company, and whether it works from your notes and past drafts every time.

That surrounding layer is a harness, also called an AI workspace. ChatGPT Business, Claude Team, Notion and EVY are all versions of it, and they differ a lot. We compared nine for small teams, EVY included, with the trade-offs spelled out. See the best AI workspaces for small teams →

Cite this ranking

Writers and researchers are welcome to quote and link this ranking. Please credit the benchmark sources named above as well.

EVY (2026). Best LLMs for writing: Claude, GPT and Gemini scored for creative writing and copywriting. https://evy.so/compare/best-llms-for-writing/ (updated October 12, 2026).

EVY

A free harness for
professional knowledge work.

EVY keeps your company's knowledge in plain files, plans your day, and hands tasks to the Claude or ChatGPT plan you already pay for. A Mac app, free.

See how EVY works → Or start with your voice: make a free voice.md

Frequently asked questions

Which LLM is best for creative writing in 2026?

OpenAI's new GPT-6 Astra leads the EQ-Bench Creative Writing leaderboard at an Elo of 2173, with Claude Fable 5.1 (Anthropic) second at 2162 and Claude Opus 5 (Anthropic) third at 2133 as of October 2026. GPT-6 Astra tops the creative benchmark, but Claude Opus 5.5 leads LMArena Text at 1507, and Claude Fable 5 leads EQ General at 2050, where Claude Opus 4.8 is a close second at 2030. These models excel at narrative quality, emotional depth, and character voice — the core skills that separate great writing from generic AI output.

What is EQ-Bench and why does it matter for writing?

EQ-Bench is an independent benchmark that evaluates large language models on emotional intelligence and narrative quality, using a panel of human raters. Its Creative Writing sub-leaderboard specifically measures story quality, emotional resonance, and prose style — making it the most relevant benchmark for marketing copy, long-form content, and creative work. Scores are on an Elo scale where higher is better, typically ranging from ~1300 to ~2175.

What is LMArena Text and how is it different from EQ-Bench?

LMArena Text (formerly LMSYS Chatbot Arena) measures human preference through head-to-head votes: two anonymous models answer the same prompt, and users pick the better response. It's a broad preference signal across all text tasks, not just writing. EQ-Bench Creative Writing is narrower and more specialist — it specifically evaluates narrative and emotional writing quality with trained raters rather than crowd votes.

Which LLM is the best value for writing tasks?

Zhipu AI's new GLM-5.3 is the standout value pick: an EQ-Bench Creative score of 2075 at just $1.40 input / $4.40 output per 1M tokens — near-frontier writing for a fraction of the premium-tier price. Meta's Muse Spark 1.1 is another strong performance-per-dollar option at 1927 Creative for $1.25 input / $4.25 output. GLM-5.2 (Zhipu AI) is another strong performance-per-dollar option at 1757 Creative for $1.40 input / $4.40 output — roughly 3.5× cheaper than Claude Opus 4.7 with ~90% of its creative writing performance. OpenAI's GPT-5.6 Luna is the deepest budget pick at 1829 Creative for just $0.20 input / $1.20 output — cheaper than Kimi K2 on both sides of the meter and 163 Elo ahead of it. Kimi K2 by Moonshot AI remains a solid low-cost option at 1666 Creative for $0.55 input / $2.20 output, and Kimi K2.6 scores even higher at 1725 Creative. GLM-5 (Zhipu AI) is another strong value option at $0.80/$2.50 with scores of 1601 EQ Creative and 1526 EQ General.

How often is this ranking updated?

Scores are updated weekly via an automated scraper that fetches the latest data from EQ-Bench and LMArena. Prices are reviewed manually and updated when providers announce changes. The 'Updated weekly' badge in the table header shows the date of the last successful update.

What does 'tokens per second' mean for writing?

Tokens per second (t/s) measures how fast a model outputs text — roughly, 75 tokens equals about 55 words. For writing workflows, speed matters when you need rapid iteration on drafts or real-time dictation-to-copy conversion. MiniMax M2.5 is the fastest tracked model at 395 t/s; Gemini 3 Flash at 250 t/s offers the best speed-to-cost ratio among paid models.

Does the best LLM for writing change depending on the task?

Yes — significantly. GPT-6 Astra tops EQ Creative (2173) with Claude Fable 5.1 right behind (2162), while Claude Fable 5 leads EQ General (2050) — making it especially strong for character voice and emotional tone, with Claude Opus 4.8 close behind at 2030. Claude Sonnet 4.6 remains excellent for structured commercial copy where format consistency matters at a lower price. GPT-5.5 is a strong contender for creative work. Faster models like Gemini 3 Flash or Grok 4.1 suit high-volume, lower-stakes content.

Does picking the best LLM guarantee good writing?

No. The model sets the ceiling on quality, but it does not decide how the writing sounds: your voice, your angle, your point of view. Output from even the top-ranked models often reads as generic and gets flagged as AI, which costs reach. The fix is to give the model your voice and your context every time. EVY keeps your company's knowledge, including how you write, in files that Claude Code and Codex read before every task, so the model you picked here starts already knowing your work.