/best/writingupdated August 2026137 models evaluated

Best AI for writing

Long-form drafting, rewriting, tone-matching, and editorial work.

CodingMathWritingResearchTranslationData AnalysisChatbotsStudentsBusinessCreative WritingTabular Data & Spreadsheets

Writing is where models diverge most on taste. Two models can score identically on instruction-following and produce prose with completely different voice — one tight and readable, the other syrupy and evasive. Benchmark scores alone won't tell you which model matches your brand.

What matters: rhetorical range, willingness to cut, instruction adherence at the paragraph level, and refusal discipline. Good models don't moralize at the drop of a hat. We weight side-by-side human preference strongly here, along with tone consistency across long documents.

Our writing rank weights writing quality (2.5×), instruction-following (1.5×), and reasoning (0.5×). For editorial work at scale, instruction-following is the main differentiator — models that silently ignore style constraints waste editorial time.

Full rankings

All 137 models, scored for writing

weighted composite · lower-is-worse
#ModelProviderTask score$/in$/outContext
01Qwen: Qwen3.7 FlashQQwen5.00/5.0$0.030$0.1301M
02OpenAI: GPT-5.6 LunaOOpenAI5.00/5.0$0.200$1.201.1M
03Gemini 3.6 FlashGGoogle5.00/5.0$0.750$3.751.0M
04OpenAI: GPT-5.6 SolOOpenAI5.00/5.0$2.00$10.001.1M
05MoonshotAI: Kimi K3MMoonshotAI5.00/5.0$3.00$15.001.0M
06OpenAI: GPT-5.6 Sol ProOOpenAI5.00/5.0$5.00$30.001.1M
07Anthropic: Claude Fable 5AAnthropic5.00/5.0$10.00$50.001M
08Claude Opus 5 (Fast)AAnthropic5.00/5.0$10.00$50.001M
09DeepSeek V4 FlashDDeepSeek4.50/5.0$0.089$0.1771.0M
10DeepSeek: DeepSeek V4 Flash 0731DDeepSeek4.50/5.0$0.140$0.2801.3M
11MiMo-V2.5XXiaomi4.50/5.0$0.140$0.2801.1M
12GLM-4.7 FlashZZhipu AI4.50/5.0$0.060$0.400203K
13NVIDIA: Nemotron 3 SuperNNVIDIA4.50/5.0$0.085$0.4001M
14MiniMax: MiniMax M3MMinimax4.50/5.0$0.300$1.201.0M
15Thinking Machines: Inkling SmallTThinkingmachines4.50/5.0$0.450$1.201.0M
16Qwen: Qwen3.7 PlusQQwen4.50/5.0$0.320$1.281M
17DeepSeek V4 ProDDeepSeek4.50/5.0$0.790$1.581.0M
18Qwen: Qwen3.6 PlusQQwen4.50/5.0$0.325$1.951M
19R1 0528DDeepSeek4.50/5.0$0.500$2.15164K
20Devstral 2 2512MMistral4.50/5.0$0.440$2.20262K
21Gemini 3.5 Flash LiteGGoogle4.50/5.0$0.300$2.501.0M
22R1DDeepSeek4.50/5.0$0.700$2.5064K
23Grok 4.3XXAI4.50/5.0$1.25$2.501M
24Gemini 3 Flash PreviewGGoogle4.50/5.0$0.500$3.001.0M
25Qwen: Qwen3.6 27BQQwen4.50/5.0$0.320$3.20262K
26Kimi K2.7 CodeMMoonshot AI4.50/5.0$0.670$3.40262K
27NVIDIA: Nemotron 3 UltraNNVIDIA4.50/5.0$0.600$3.60512K
28Z.ai: GLM 5.2ZZhipu AI4.50/5.0$1.19$3.741.0M
29MoonshotAI: Kimi K2.6MMoonshotAI4.50/5.0$0.950$4.00262K
30Meta: Muse Spark 1.1MMeta4.50/5.0$1.25$4.251.0M
31Qwen 3.7 MaxQQwen4.50/5.0$1.48$4.421M
32OpenAI: GPT-5.6 Luna ProOOpenAI4.50/5.0$1.00$6.001.1M
33Qwen: Qwen3.6 Max PreviewQQwen4.50/5.0$1.03$6.16262K
34Grok 4.5XXAI4.50/5.0$2.00$6.00500K
35Gemini 3.5 FlashGGoogle4.50/5.0$1.50$9.001.0M
36Anthropic: Claude Sonnet 5AAnthropic4.50/5.0$2.00$10.001M
37Gemini 3.1 Pro PreviewGGoogle4.50/5.0$2.00$12.001.0M
38OpenAI: GPT-5.6 TerraOOpenAI4.50/5.0$2.00$12.001.1M
39GPT-5.2OOpenAI4.50/5.0$1.75$14.00400K
40OpenAI: GPT-5.6 Terra ProOOpenAI4.50/5.0$2.50$15.001.1M
41Claude Opus 5AAnthropic4.50/5.0$5.00$25.001M
42Claude Opus 4.7AAnthropic4.50/5.0$5.00$25.001M
43GPT-5.5OOpenAI4.50/5.0$5.00$30.001.1M
44Anthropic: Claude Opus 4.8 (Fast)AAnthropic4.50/5.0$10.00$50.001M
45Ministral 3 3B 2512MMistral4.00/5.0$0.100$0.100131K
46OpenAI: gpt-oss-120bOOpenAI4.00/5.0$0.037$0.170131K
47Qwen: Qwen3.5-9BQQwen4.00/5.0$0.100$0.150262K
48Ministral 3 8B 2512MMistral4.00/5.0$0.150$0.150262K
49Qwen: Qwen3 30B A3B Instruct 2507QQwen4.00/5.0$0.048$0.193262K
50Ministral 3 14B 2512MMistral4.00/5.0$0.200$0.200262K
51Gemma 4 31BGGoogle4.00/5.0$0.100$0.340262K
52DeepSeek V3.2DDeepSeek4.00/5.0$0.260$0.380164K
53Grok 4.1 FastXXAI4.00/5.0$0.200$0.5002M
54Qwen: Qwen3 235B A22B Instruct 2507QQwen4.00/5.0$0.090$0.550262K
55inclusionAI: Ling-2.6-1TIInclusionai4.00/5.0$0.075$0.625262K
56Arcee AI: Trinity Large ThinkingAArcee-ai4.00/5.0$0.220$0.850262K
57MiMo-V2.5-ProXXiaomi4.00/5.0$0.435$0.8701.1M
58Qwen: Qwen3.6 35B A3BQQwen4.00/5.0$0.140$1.00262K
59Nex-N2-ProNNex AGI4.00/5.0$0.250$1.00262K
60MiniMax: MiniMax M2MMinimax4.00/5.0$0.255$1.02205K
61Qwen: Qwen3.6 FlashQQwen4.00/5.0$0.188$1.131M
62GPT-5.4 NanoOOpenAI4.00/5.0$0.200$1.25400K
63Qwen: Qwen3.5-35B-A3BQQwen4.00/5.0$0.250$1.25262K
64Gemini 3.1 Flash Lite PreviewGGoogle4.00/5.0$0.250$1.501.0M
65Google: Gemini 3.1 Flash LiteGGoogle4.00/5.0$0.250$1.501.0M
66DeepSeek V3.1DDeepSeek4.00/5.0$0.550$1.65164K
67Qwen: Qwen3.5 Plus 2026-04-20QQwen4.00/5.0$0.300$1.801M
68GPT-5 MiniOOpenAI4.00/5.0$0.250$2.00400K
69Mistral Medium 3.1MMistral4.00/5.0$0.400$2.00131K
70xAI: Grok Build 0.1XXAI4.00/5.0$1.00$2.00256K
71Gemini 2.5 FlashGGoogle4.00/5.0$0.300$2.501.0M
72Grok 4.20XXAI4.00/5.0$1.25$2.502M
73Thinking Machines: InklingTThinkingmachines4.00/5.0$0.950$4.051.0M
74GPT-5.4 MiniOOpenAI4.00/5.0$0.750$4.50400K
75o4 MiniOOpenAI4.00/5.0$1.10$4.40200K
76Grok 4.6XXAI4.00/5.0$2.00$6.00500K
77Mistral Medium 3.5MMistral4.00/5.0$1.50$7.50262K
78GPT-4.1OOpenAI4.00/5.0$2.00$8.001.0M
79o3OOpenAI4.00/5.0$2.00$8.00200K
80GPT-5OOpenAI4.00/5.0$1.25$10.00400K
81GPT-5.1OOpenAI4.00/5.0$1.25$10.00400K
82Gemini 2.5 ProGGoogle4.00/5.0$1.25$10.001.0M
83GPT-5.4OOpenAI4.00/5.0$2.50$15.001.1M
84Claude Sonnet 4.6AAnthropic4.00/5.0$3.00$15.001M
85Anthropic: Claude Opus 4.8AAnthropic4.00/5.0$5.00$25.001M
86Claude Opus 4.6AAnthropic4.00/5.0$5.00$25.001M
87inclusionAI: Ling-2.6-flashIInclusionai3.50/5.0$0.010$0.030262K
88IBM: Granite 4.1 8BIIbm-granite3.50/5.0$0.050$0.100131K
89Poolside: Laguna XS 2.1PPoolside3.50/5.0$0.060$0.120262K
90Xiaomi: MiMo-V2-FlashXXiaomi3.50/5.0$0.100$0.300262K
91StepFun: Step 3.5 FlashSStepfun3.50/5.0$0.100$0.300262K
92Gemma 4 26B A4B GGoogle3.50/5.0$0.070$0.340262K
93ByteDance Seed: Seed-2.0-MiniBBytedance-seed3.50/5.0$0.100$0.400262K
94Gemini 2.5 Flash LiteGGoogle3.50/5.0$0.100$0.4001.0M
95Grok 3 MiniXXAI3.50/5.0$0.300$0.500131K
96Mistral Small 4MMistral3.50/5.0$0.150$0.600262K
97inclusionAI: Ring-2.6-1TIInclusionai3.50/5.0$0.075$0.625262K
98Inception: Mercury 2IInception3.50/5.0$0.250$0.750128K
99DeepSeek V3.1 TerminusDDeepSeek3.50/5.0$0.270$1.00164K
100MiniMax: MiniMax M2.5MMinimax3.50/5.0$0.270$1.08205K
101StepFun: Step 3.7 FlashSStepfun3.50/5.0$0.200$1.15262K
102Kwaipilot: KAT-Coder-Pro V2KKwaipilot3.50/5.0$0.300$1.20262K
103MiniMax: MiniMax M2.1MMinimax3.50/5.0$0.300$1.20205K
104MiniMax: MiniMax M2.7MMinimax3.50/5.0$0.300$1.20205K
105GPT-4.1 MiniOOpenAI3.50/5.0$0.400$1.601.0M
106GLM-4.7ZZhipu AI3.50/5.0$0.400$1.75205K
107ByteDance Seed: Seed-2.0-LiteBBytedance-seed3.50/5.0$0.250$2.00262K
108ByteDance Seed: Seed 1.6BBytedance-seed3.50/5.0$0.250$2.00262K
109ByteDance Seed: Seed-2.0-CodeBBytedance-seed3.50/5.0$0.500$3.00262K
110Claude Haiku 4.5AAnthropic3.50/5.0$1.00$5.00200K
111Grok 4XXAI3.50/5.0$3.00$15.00256K
112OpenAI: gpt-oss-20bOOpenAI3.00/5.0$0.030$0.130131K
113Poolside Laguna S 2.1PPoolside3.00/5.0$0.090$0.1801.0M
114Mistral Small 3.2 24BMMistral3.00/5.0$0.075$0.200131K
115ByteDance Seed: Seed 1.6 FlashBBytedance-seed3.00/5.0$0.075$0.300262K
116Llama 4 ScoutMMeta3.00/5.0$0.100$0.3001.3M
117Llama 3.3 70B InstructMMeta3.00/5.0$0.100$0.320131K
118GPT-5 NanoOOpenAI3.00/5.0$0.050$0.400400K
119GPT-4.1 NanoOOpenAI3.00/5.0$0.100$0.4001.0M
120KAT-Coder-Air V2.5KKwaipilot3.00/5.0$0.150$0.600256K
121Llama 4 MaverickMMeta3.00/5.0$0.200$0.8001.0M
122MiniMax: MiniMax-01MMinimax3.00/5.0$0.200$1.101.0M
123Grok Code Fast 1XXAI3.00/5.0$0.200$1.50256K
124Mistral Large 3 2512MMistral3.00/5.0$0.500$1.50262K
125MiniMax: MiniMax M1MMinimax3.00/5.0$0.550$2.201M
126GPT-4oOOpenAI3.00/5.0$2.50$10.00128K
127Grok 3XXAI3.00/5.0$3.00$15.00131K
128IBM: Granite 4.0 MicroIIbm-granite2.50/5.0$0.017$0.112131K
129NVIDIA: Nemotron 3 Nano 30B A3BNNVIDIA2.50/5.0$0.050$0.200262K
130Qwen: Qwen3 Coder 30B A3B InstructQQwen2.50/5.0$0.070$0.280262K
131GPT-4o-miniOOpenAI2.50/5.0$0.150$0.600128K
132Mistral Small 3.1 24BMMistral2.50/5.0$0.351$0.555128K
133Codestral 2508MMistral2.50/5.0$0.300$0.900256K
134Seed 2.1 TurboBByteDance2.00/5.0$0.500$2.50262K
135MiniMax: MiniMax M2-herMMinimax1.50/5.0$0.300$1.2066K
136KAT-Coder-Pro V2.5KKwaipilot0.50/5.0$0.740$2.96256K
137Qwen: Qwen3.8 MaxQQwen0.00/5.0$2.00$6.001M

Pricing — top 5 for writing

QQwen: Qwen3.7 Flash
$0.105/MTok
5.00/5.0
OOpenAI: GPT-5.6 Luna
$0.950/MTok
5.00/5.0
GGemini 3.6 Flash
$3.00/MTok
5.00/5.0
OOpenAI: GPT-5.6 Sol
$8.00/MTok
5.00/5.0
MMoonshotAI: Kimi K3
$12.00/MTok
5.00/5.0
modelpicker.aipowered by live benchmark data

The best AI for writing changes every month.

We'll email you when rankings shift, new models hit the top 5, or pricing cuts reshuffle the value leaders.

Get notified when models change
Price drops, new models, benchmark updates. One email per change, no spam.