Writing

🏆 Leaderboard

As of August 21, 2026, Claude Fable 5 is #1 for writing at 1506. Ranked by Chatbot Arena human votes: people pick the better reply in a blind test. That is a writing and taste rank, not a trivia quiz. Best AI for writing, ranked by Chatbot Arena votes (humans pick the better reply, blind). MMLU-Pro is a knowledge quiz and is not how this page picks a winner.

Updated August 21, 2026282 models33 providers


Claude Fable 5 Anthropic · Proprietary	1506	91.5%	—	68.3%	78.3%	—	—	—	—	—	—	$60.00
Muse Spark 1.2 Meta · Proprietary	1498	88.3%	—	—	—	—	—	—	—	—	—	$5.50
Claude Opus 4.6 Anthropic · Proprietary	1497	—	—	41.0%	76.3%	—	91.1%	—	—	—	—	$30.00
Claude Opus 4.7 Anthropic · Proprietary	1494	89.9%	—	50.6%	76.9%	—	91.5%	—	—	—	—	$30.00
Claude Opus 5 Anthropic · Proprietary	1493	91.6%	—	56.7%	—	—	—	—	—	—	—	$30.00
Qwen3.8 MaxOSS Qwen · Open Source	1491	88.6%	—	46.3%	—	—	—	—	82.8%	—	—	$8.00
Gemini 3.7 Flash Google · Proprietary	1490	90.1%	—	71.2%	—	—	—	—	—	—	—	$4.50
Muse Spark 1.1 Meta · Proprietary	1489	88.7%	—	—	—	—	—	—	—	—	—	$5.50
Kimi K3OSS Moonshot AI · Open Source	1489	88.0%	—	42.7%	—	—	—	—	—	—	—	$18.00
Gemini 3.1 Pro Google · Proprietary	1486	91.0%	—	77.3%	79.9%	—	92.6%	—	—	—	—	$17.50
Gemini 3 Pro Google · Proprietary	1485	90.1%	—	72.1%	73.4%	—	91.8%	—	—	—	—	$14.00
Gemini 3.6 Flash Google · Proprietary	1484	89.3%	—	68.7%	—	—	—	—	—	—	—	$4.50
GPT-5.5 OpenAI · Proprietary	1482	88.1%	—	—	80.7%	—	—	—	—	—	—	$35.00
GPT-5.6 Sol OpenAI · Proprietary	1481	89.1%	—	71.6%	—	—	—	—	—	—	—	$35.00
Claude Opus 4.8 Anthropic · Proprietary	1481	89.6%	—	39.5%	77.2%	—	—	—	—	—	—	$30.00
Gemini 3.5 Flash Google · Proprietary	1477	89.5%	—	68.4%	75.0%	—	—	—	—	—	—	$10.50
ChatGPT-4o Latest OpenAI · Proprietary	—	—	—	—	—	—	—	—	—	—	—	$12.50
Claude 3 Haiku Anthropic · Proprietary	—	—	75.2%	—	—	—	—	—	—	85.9%	74.2%	$1.50
Claude 3 Opus Anthropic · Proprietary	—	68.5%	86.8%	—	49.2%	—	—	—	—	95.4%	88.5%	$90.00
Claude 3 Sonnet Anthropic · Proprietary	—	56.8%	79%	—	—	—	—	—	—	89%	75.1%	$18.00
Claude 3.5 Haiku Anthropic · Proprietary	—	65%	74.3%	6.7%	43.5%	—	—	7.3%	—	—	—	$4.80
Claude 3.5 Sonnet Anthropic · Proprietary	—	77.6%	90.4%	—	59.0%	—	—	8.0%	—	—	—	$18.00
Claude 3.5 Sonnet Anthropic · Proprietary	—	76.1%	90.4%	—	—	—	—	—	—	—	—	$18.00
Claude 3.7 Sonnet Anthropic · Proprietary	—	80.7%	—	—	76.1%	93.2%	86.1%	8.1%	—	—	—	$18.00
Claude Haiku 4.5 Anthropic · Proprietary	—	—	—	5.9%	—	—	83%	—	—	—	—	$6.00
Claude Mythos 5 Anthropic · Proprietary	—	—	—	—	—	—	—	—	—	—	—	$60.00
Claude Mythos Preview Anthropic · Proprietary	—	—	—	—	—	—	92.7%	—	—	—	—	$60.00
Claude Opus 4 Anthropic · Proprietary	—	86.2%	—	—	—	—	88.8%	8.4%	—	—	—	$90.00
Claude Opus 4.1 Anthropic · Proprietary	—	87.2%	—	34.8%	—	—	89.5%	8.5%	—	—	—	$90.00
Claude Opus 4.5 Anthropic · Proprietary	—	85.6%	—	41.8%	76.0%	—	90.8%	—	—	—	—	$30.00
Claude Sonnet 4 Anthropic · Proprietary	—	79.4%	—	—	—	—	86.5%	8.1%	—	—	—	$18.00
Claude Sonnet 4.5 Anthropic · Proprietary	—	—	—	23.6%	—	—	89.1%	—	—	—	—	$18.00
Claude Sonnet 4.6 Anthropic · Proprietary	—	87.3%	89.3%	29%	75.5%	—	89.3%	—	—	—	—	$18.00
Claude Sonnet 5 Anthropic · Proprietary	—	87.5%	—	25%	—	—	—	—	—	—	—	$12.00
Command A+OSS Cohere · Open Source	—	—	—	—	—	—	—	—	74%	—	—	$12.50
Command R+OSS Cohere · Open Source	—	—	75.7%	—	—	—	—	—	—	88.6%	85.4%	$1.25
Composer 2 Cursor · Proprietary	—	—	—	—	—	—	—	—	—	—	—	$3.00
Composer 2 Fast Cursor · Proprietary	—	—	—	—	—	—	—	—	—	—	—	$9.00
DeepSeek R1 Distill Llama 70BOSS DeepSeek · Open Source	—	—	—	—	54.5%	—	—	—	—	—	—	$0.50
DeepSeek R1 Distill Qwen 32BOSS DeepSeek · Open Source	—	—	—	—	45.5%	—	—	—	—	—	—	$0.30
DeepSeek-R1OSS DeepSeek · Open Source	—	83.2%	—	—	71.6%	—	—	8.3%	—	—	—	$2.74
DeepSeek-R1-0528OSS DeepSeek · Open Source	—	85%	—	92.3%	—	—	—	8.2%	—	—	—	$2.74
DeepSeek-V2.5OSS DeepSeek · Open Source	—	—	80.4%	—	—	—	—	—	—	—	—	$0.42
DeepSeek-V3OSS DeepSeek · Open Source	—	75.9%	88.5%	24.9%	60.5%	86.1%	—	—	—	88.9%	85.2%	$1.37
DeepSeek-V3 0324OSS DeepSeek · Open Source	—	81.2%	—	—	66.9%	—	—	7.7%	—	—	—	$1.42
DeepSeek-V3.1OSS DeepSeek · Open Source	—	83.7%	—	93.4%	—	—	—	—	—	—	—	$1.27
DeepSeek-V3.2OSS DeepSeek · Open Source · via OpenRouter	—	85%	—	—	—	—	—	—	—	—	—	$0.57
DeepSeek-V3.2 (Non-thinking)OSS DeepSeek · Open Source	—	—	—	—	—	—	—	—	—	—	—	$0.70
DeepSeek-V3.2-ExpOSS DeepSeek · Open Source	—	85%	—	97.1%	—	—	—	—	—	—	—	$0.68
DeepSeek-V4-Flash-0423OSS DeepSeek · Open Source	—	86.4%	—	28.9%	—	—	—	—	—	—	—	$0.30

Showing 1–50 of 282 models

Next step

You found the model. Now ship the product.

Auth, billing, and the API layer are already decided. Fork 8 finished AI apps, or have us build the first version with you.

Build with the APIs Have us build it with you

The index

All Large Language Models

282 models across 33 providers. Search or jump to a lab — every model page stays linked here.

Cursor

2 models

Composer 2 Composer 2 Fast

xAI

15 models

Grok 4.6 Grok 4.5 Grok 4.3 Grok-4.1 Grok-4.1 Fast Non-Reasoning Grok-4.1 Fast Reasoning Grok-4.1 Thinking Grok 4 Fast Grok Code Fast 1 Grok-4 Fast Non-Reasoning Grok-4 Fast Reasoning Grok-4 Grok-3 Grok-3 Mini Grok-2

DeepSeek

16 models

DeepSeek-V4-Pro-0813 DeepSeek-V4-Flash-0731 DeepSeek-V4-Flash-0423 DeepSeek-V4-Flash-Max DeepSeek-V4-Pro-Max DeepSeek-V3.2 DeepSeek-V3.2 (Non-thinking)DeepSeek-V3.2-Exp DeepSeek-R1-0528 DeepSeek-V3 0324 DeepSeek R1 Distill Llama 70B DeepSeek R1 Distill Qwen 32B DeepSeek-R1 DeepSeek-V3.1 DeepSeek-V3 DeepSeek-V2.5

MiniMax

6 models

MiniMax M3 MiniMax M2.7 MiniMax M2.5 MiniMax M2.1 MiniMax M2 MiniMax M1 80K

AI21 Labs

2 models

Jamba 1.5 Large Jamba 1.5 Mini

Amazon

5 models

Nova 2 Lite Nova 2 Sonic Nova Lite Nova Micro Nova Pro

Baidu

1 models

ERNIE 4.5

ByteDance

2 models

Seed 2.1 Turbo Seed 2.0 Pro

Cohere

3 models

North Mini Code 1.0 Command A+Command R+

IBM

1 models

Granite 3.3 8B Instruct

Inception

1 models

Mercury 2

inclusionAI

1 models

Ling-3.0-flash

LG AI Research

1 models

K-EXAONE-236B-A23B

Meituan

4 models

LongCat-Flash-Lite LongCat-Flash-Thinking-2601 LongCat-Flash-Thinking LongCat-Flash-Chat

Microsoft

5 models

MAI-Code-1.1-Flash Phi 4 Mini Phi-4-multimodal-instruct Phi 4 Phi-3.5-mini-instruct

Moonshot AI

9 models

Kimi K3 Kimi K2.7 Code Kimi K2.7 Code HighSpeed Kimi K2.6 Kimi K2.5 Kimi K2 0905 Kimi K2-Instruct-0905 Kimi K2-Thinking-0905 Kimi K2 Instruct

Nous Research

1 models

Hermes 3 70B

NVIDIA

5 models

Nemotron 3.5 Lightning (30B A3B)Nemotron 3 Ultra (550B A55B)Nemotron 3 Super (120B A12B)Nemotron 3 Nano (30B A3B)Nemotron Nano 9B v2

Poolside

2 models

Laguna S 2.1 Laguna XS 2.1

Sakana AI

1 models

Sakana Namazu

StepFun

1 models

Step-3.5-Flash

Tencent

1 models

Hy3

Thinking Machines

1 models

Inkling-Small

Unisound

1 models

Upstage

1 models

Solar Pro 4

Xiaomi

3 models

MiMo-V2.5-Pro MiMo-V2.5 MiMo-V2-Flash

What is the best AI for writing?

As of August 21, 2026, Claude Fable 5 by Anthropic is #1 for writing at 1506. Ranked by Chatbot Arena human votes: people pick the better reply in a blind test. That is a writing and taste rank, not a trivia quiz. This board also tracks Chatbot Arena, MMLU-Pro, MMLU, SimpleQA. Next on the same board: Muse Spark 1.2 and Claude Opus 4.6. Related leaders: Claude Opus 5 on MMLU-Pro at 91.6%; GPT-5 on MMLU at 92.5%. This writing leaderboard ranks models by Chatbot Arena. Scores come from public evals. Prices are the live API rates in the table above.

Sources: Chatbot Arena (LMArena) (https://lmarena.ai/); OpenAI API pricing (https://developers.openai.com/api/docs/pricing); Anthropic Claude API pricing (https://platform.claude.com/docs/en/about-claude/pricing); Gemini API pricing (https://ai.google.dev/gemini-api/docs/pricing)

Top 8 for writing. Ranked by Chatbot Arena human votes: people pick the better reply in a blind test. That is a writing and taste rank, not a trivia quiz. Input and output are dollars per million tokens.
Rank	Model	Chatbot Arena	Input /M	Output /M
1	Claude Fable 5	1506	$10.00	$50.00
2	Muse Spark 1.2	1498	$1.25	$4.25
3	Claude Opus 4.6	1497	$5.00	$25.00
4	Claude Opus 4.7	1494	$5.00	$25.00
5	Claude Opus 5	1493	$5.00	$25.00
6	Qwen3.8 Max	1491	$2.00	$6.00
7	Gemini 3.7 Flash	1490	$0.75	$3.75
8	Kimi K3	1489	$3.00	$15.00

What is MMLU-Pro?

MMLU-Pro is a multiple-choice knowledge exam: law, science, math, and similar. People see it on leaderboards because it is dense, not because it measures writing. A model can ace MMLU-Pro and still write like a press release. This page ranks writing by Chatbot Arena votes: two replies, no names, humans pick a winner.

Best AI writer vs best writing model

Search results for best AI for writing are full of app roundups. Those tools wrap a model. This leaderboard ranks the models on preference and language evals, with token price. If you already pay for ChatGPT or Claude, you already have a writer. The question is which SKU to call in production.

Claude vs ChatGPT for writing

Taste tests usually go to Claude. Tight formats and tool-using drafts often go to GPT. MMLU will not tell you that. Sort Arena Elo, read SimpleQA so you do not ship a fluent liar, then check $/M if you generate in bulk.

Related LLM leaderboards

LLM Leaderboard Coding Math Agentic Reasoning Science Finance Long Context Vision Open Source

Individual benchmark pages

Rank one eval at a time. All LLM benchmarks.

HLE SWE Bench Humanity's Last Exam (No Tools)Humanity's Last Exam (With Tools)Terminal-Bench 2.0 Terminal-Bench Terminal-Bench 2.1 Terminal-Bench 2 Terminal-Bench 3.0 ARC-AGI-1 Verified

Writing FAQ

Who ranks #1 on the Writing leaderboard?

As of August 21, 2026, Claude Fable 5 by Anthropic ranks #1 on Chatbot Arena at 1506. API pricing is $10.00/M input and $50.00/M output.

What is the best AI for writing?

What are the top models on Chatbot Arena?

The current Chatbot Arena ranking as of August 21, 2026 is 1. Claude Fable 5 at 1506; 2. Muse Spark 1.2 at 1498; 3. Claude Opus 4.6 at 1497.

Which writing model is the cheapest?

Gemini 3.7 Flash is the cheapest scored model on this writing leaderboard at $0.75/M input and $3.75/M output ($4.50 blended). Claude Fable 5 still leads Chatbot Arena at 1506.

Should I always pick the #1 Chatbot Arena model?

Not automatically. Claude Fable 5 leads Chatbot Arena, but a cheaper scored model can be the better production choice if the quality gap is small. Use the table to weigh Chatbot Arena against input/output price, context window, and related evals.

How often is the Writing leaderboard updated?

Scores and API prices on this page are refreshed from published evals and provider rates. The snapshot is labeled August 21, 2026. Treat it as a current index, not a one-off blog post.

What is the best LLM for writing?

This page ranks writing by Chatbot Arena votes: humans pick the better reply without seeing the names. That is a taste and prose signal. MMLU-Pro is a multiple-choice knowledge quiz and is not the ranking.

What is MMLU-Pro?

MMLU-Pro is a hard multiple-choice exam (science, law, math, and similar). A high MMLU-Pro score means the model knows facts. It does not mean the model writes well. The writing board still shows the column so you can sort it, but the default rank is Arena votes.

Does MMLU measure writing style?

No. MMLU and MMLU-Pro measure knowledge, not prose. HellaSwag is sentence completion. For writing, use Chatbot Arena votes or the Lech Mazur writing eval, then read a sample yourself.

Should I pick the cheapest writing model?

Only if the writing is good enough. Rank by Arena votes first, then check token price. A cheap model that sounds stiff still costs you edits.

What is the best AI writer?

Best AI writer usually means a writing app, not a model. This page ranks the models those apps call by Chatbot Arena votes (humans pick the better reply, blind), then shows API price. MMLU-Pro is a knowledge quiz and is not the ranking. Jasper vs ChatGPT is a product comparison.

What is the best AI for creative writing?

Creative writing is preference-heavy. Arena Elo and writing-specific evals beat MMLU. Claude often wins taste tests. GPT often wins instruction-following. Sample both. Do not trust a knowledge exam for prose.

What is the best AI for essays?

Essay help is a mix of structure, citation, and hallucination rate. SimpleQA is the factuality check. MMLU is knowledge, not voice. Rank this board, then read a page of output yourself.

Claude vs ChatGPT for writing: which is better?

Claude vs ChatGPT writing is the usual split: Claude for tone, GPT for following a tight brief. This table puts preference scores and price on one row so you can stop arguing from anecdotes.

What is MMLU-Pro, and why isn't writing ranked by it?

MMLU-Pro is a hard multiple-choice knowledge test. Labs love it because almost every model has a score. It says nothing about whether the prose is any good. This writing board defaults to Chatbot Arena votes. You can still sort the MMLU-Pro column if you want a quiz rank.

Does MMLU measure writing quality?

No. MMLU is a knowledge test. HellaSwag is sentence completion. Neither is a prose judge. Use Arena Elo and a writing eval, then your own samples.

You found the model. Now ship the product.

All Large Language Models

OpenAI

Anthropic

Google

Cursor

xAI

DeepSeek

MiniMax

Qwen

AI21 Labs

Amazon

Baidu

ByteDance

Cohere

IBM

Inception

inclusionAI

LG AI Research

Meituan

Meta

Microsoft

Mistral

Moonshot AI

Nous Research

NVIDIA

Poolside

Sakana AI

StepFun

Tencent

Thinking Machines

Unisound

Upstage

Xiaomi

Z AI

What is the best AI for writing?

What is MMLU-Pro?

Best AI writer vs best writing model

Claude vs ChatGPT for writing

Related LLM leaderboards

Individual benchmark pages

Writing FAQ

Who ranks #1 on the Writing leaderboard?

What is the best AI for writing?

What are the top models on Chatbot Arena?

Which writing model is the cheapest?

Should I always pick the #1 Chatbot Arena model?

How often is the Writing leaderboard updated?

What is the best LLM for writing?

What is MMLU-Pro?

Does MMLU measure writing style?

Should I pick the cheapest writing model?

What is the best AI writer?

What is the best AI for creative writing?

What is the best AI for essays?

Claude vs ChatGPT for writing: which is better?

What is MMLU-Pro, and why isn't writing ranked by it?

Does MMLU measure writing quality?