Romance: separated couple reunites

Bad Writing Habits

Detects common prose quality anti-patterns in AI-generated creative writing, including passive voice, past progressive overuse, weak dialogue tags, filter words, purple prose, cliches, AI-ism words/adverbs/names, and more.

Creative Writing Hallucination

Performance Score Distribution (Top 20)

Click a model name to view its detail page.

	Score
Claude Haiku 5.5 (Reasoning, Low)	96%
GPT-5.4 (Reasoning, Low)	92%
GPT-5.4	92%
Z.AI GLM 5.3 (Reasoning, Max)	91%
GPT-6.1 Sol (Reasoning, Medium)	91%
DeepSeek V4.1 Flash (Reasoning, High)	90%
Writer: Palmyra X5	90%
Claude Opus 5.5 (Reasoning)	90%
Aion 3.5 (Reasoning, High)	90%
Qwen 3.7 Flash (Reasoning)	90%
Xiaomi MIMO v2.6 Pro	90%
Z.AI GLM 5.3 Flash (Reasoning, Max)	90%
Claude Sonnet 5.5 (Reasoning)	90%
Grok 4.5 (Reasoning, Low)	90%
Z.AI GLM 5.2 (Reasoning, High)	90%
GPT-5.6 Terra	89%
GPT-5.5	89%
Claude Opus 5 (Reasoning)	89%
GPT-5.6 Luna	89%
Claude Opus 4.6 (Reasoning)	89%

	Score	Cost	Time
Claude Haiku 5.5 (Reasoning, Low)	96%	$0.0021	18.9s
Qwen 3.7 Flash (Reasoning)	90%	$0.0012	1.0m
Z.AI GLM 5.3 Flash (Reasoning, Low)	88%	$0.0006	1.0m
DeepSeek V4.1 Flash (Reasoning, High)	90%	$0.0040	1.1m
Writer: Palmyra X5	90%	$0.013	22.8s
Z.AI GLM 5.2 (Reasoning, High)	90%	$0.012	57.5s
Grok 4.5 (Reasoning, Low)	90%	$0.019	50.7s
GPT-5.6 Luna	89%	$0.016	18.9s
GPT-6 Luna (Reasoning, Medium)	89%	$0.0014	36.3s
GPT-5.4 (Reasoning, Low)	92%	$0.056	1.3m
Muse Spark 1.1 (Reasoning, Medium)	88%	$0.016	22.4s
GPT-5.4 Mini	89%	$0.014	15.7s
Claude Haiku 5.5 (Reasoning, Medium)	88%	$0.0017	14.1s
GPT-5.4 Mini (Reasoning, Low)	88%	$0.014	16.1s
GPT-5.6 Luna (Reasoning, Medium)	89%	$0.017	19.0s
GPT-6.1 Sol (Reasoning, Medium)	91%	$0.033	1.6m
Gemini 3.5 Flash Lite (Reasoning, Minimal)	87%	$0.0041	6.1s
GPT-5.4	92%	$0.051	1.3m
GPT-5.6 Terra	89%	$0.038	28.7s
DeepSeek V4 Flash (Reasoning)	88%	$0.0007	25.0s

	Score	Consistency	Stability
GPT-5.4 (Reasoning, Low)	92%	97%	89%
Claude Sonnet 5.5 (Reasoning)	90%	98%	88%
GPT-5.4	92%	95%	88%
Claude Opus 5.5 (Reasoning)	90%	97%	88%
GPT-5.6 Sol	88%	99%	88%
GPT-6.1 Sol (Reasoning, Medium)	91%	96%	88%
Grok 4.7 (Reasoning, High)	88%	98%	87%
GPT-5.6 Terra	89%	98%	87%
GPT-5.5 (Reasoning, Medium)	89%	98%	87%
DeepSeek V4.1 Flash (Reasoning, High)	90%	96%	87%
DeepSeek V4 Pro	88%	99%	87%
GPT-6 Luna (Reasoning, Medium)	89%	98%	87%
Z.AI GLM 5.3 Flash (Reasoning, Max)	90%	94%	87%
GPT-5.5 (Reasoning, Low)	88%	97%	87%
Grok 4.6 (Reasoning, High)	89%	96%	86%
Xiaomi MIMO v2.6 Pro	90%	96%	86%
Qwen 3.7 Flash (Reasoning)	90%	96%	86%
GPT-5.6 Luna (Reasoning, Medium)	89%	98%	86%
GPT-5.5	89%	97%	86%
GPT-5.6 Luna	89%	97%	86%

	Score	Cost	Speed	Stability
Claude Haiku 5.5 (Reasoning, Low)	96%	$0.0021	18.9s	—
GPT-5.4 (Reasoning, Low)	92%	$0.056	1.3m	89%
Writer: Palmyra X5	90%	$0.013	22.8s	85%
DeepSeek V4.1 Flash (Reasoning, High)	90%	$0.0040	1.1m	87%
GPT-5.4 Mini	89%	$0.014	15.7s	86%
GPT-6 Luna (Reasoning, Medium)	89%	$0.0014	36.3s	87%
GPT-5.6 Luna	89%	$0.016	18.9s	86%
Qwen 3.7 Flash (Reasoning)	90%	$0.0012	1.0m	86%
GPT-5.6 Luna (Reasoning, Medium)	89%	$0.017	19.0s	86%
GPT-5.6 Terra	89%	$0.038	28.7s	87%
GPT-5.4	92%	$0.051	1.3m	88%
Claude Sonnet 5.5 (Reasoning)	90%	$0.053	41.0s	88%
Muse Spark 1.1 (Reasoning, Medium)	88%	$0.016	22.4s	86%
GPT-5.4 Mini (Reasoning, Low)	88%	$0.014	16.1s	85%
Xiaomi MIMO v2.6 Pro	90%	$0.0026	1.6m	86%
GPT-6.1 Sol (Reasoning, Medium)	91%	$0.033	1.6m	88%
DeepSeek V4 Pro	88%	$0.0027	1.1m	87%
Aion 3.5 (Reasoning, High)	90%	$0.023	1.3m	86%
DeepSeek V4 Flash (Reasoning)	88%	$0.0007	25.0s	83%
Z.AI GLM 5.2 (Reasoning, High)	90%	$0.012	57.5s	84%

Rank	Model	Avg. Cost	Avg. Time	Stability	# 1	# 2	# 3	# 4	# 5	Total
1	Claude Haiku 5.5 (Reasoning, Low)	$0.0021	18.9s	–	96	–	–	–	–	96%
2	GPT-5.4 (Reasoning, Low)	$0.056	1.3m	89%	95	94	93	91	90	92%
11	GPT-5.4	$0.051	1.3m	88%	95	93	92	90	88	92%
180	Z.AI GLM 5.3 (Reasoning, Max)	$0.086	8.8m	85%	95	93	90	89	87	91%
16	GPT-6.1 Sol (Reasoning, Medium)	$0.033	1.6m	88%	93	92	91	91	87	91%
4	DeepSeek V4.1 Flash (Reasoning, High)	$0.0040	1.1m	87%	92	92	91	90	86	90%
3	Writer: Palmyra X5	$0.013	22.8s	85%	94	91	90	89	86	90%
30	Claude Opus 5.5 (Reasoning)	$0.108	56.4s	88%	92	91	90	89	89	90%
18	Aion 3.5 (Reasoning, High)	$0.023	1.3m	86%	93	91	89	89	88	90%
8	Qwen 3.7 Flash (Reasoning)	$0.0012	1.0m	86%	92	92	90	88	87	90%
15	Xiaomi MIMO v2.6 Pro	$0.0026	1.6m	86%	93	91	90	89	87	90%
163	Z.AI GLM 5.3 Flash (Reasoning, Max)	$0.0049	9.4m	87%	93	92	92	87	86	90%
12	Claude Sonnet 5.5 (Reasoning)	$0.053	41.0s	88%	92	90	90	89	89	90%
29	Grok 4.5 (Reasoning, Low)	$0.019	50.7s	82%	96	93	90	87	82	90%
20	Z.AI GLM 5.2 (Reasoning, High)	$0.012	57.5s	84%	93	93	90	87	85	90%
10	GPT-5.6 Terra	$0.038	28.7s	87%	92	89	89	89	88	89%
53	GPT-5.5	$0.118	1.5m	86%	92	90	89	88	88	89%
56	Claude Opus 5 (Reasoning)	$0.107	56.6s	84%	93	91	88	88	86	89%
7	GPT-5.6 Luna	$0.016	18.9s	86%	91	90	89	88	86	89%
49	Claude Opus 4.6 (Reasoning)	$0.098	1.3m	85%	91	91	89	89	85	89%
32	GPT-6 Sol (Reasoning, Medium)	$0.037	1.1m	84%	93	90	89	88	85	89%
24	GPT-5.6 Terra (Reasoning, Medium)	$0.043	32.8s	85%	92	89	88	88	86	89%
57	GPT-5.6 Sol (Reasoning, Medium)	$0.115	1.2m	85%	92	89	88	88	87	89%
5	GPT-5.4 Mini	$0.014	15.7s	86%	91	89	88	88	87	89%
74	Grok 4.6 (Reasoning, High)	$0.061	3.4m	86%	91	90	90	87	86	89%
71	GPT-5.5 (Reasoning, Medium)	$0.140	1.7m	87%	90	89	89	88	87	89%
6	GPT-6 Luna (Reasoning, Medium)	$0.0014	36.3s	87%	90	89	89	87	87	89%
9	GPT-5.6 Luna (Reasoning, Medium)	$0.017	19.0s	86%	91	89	88	88	87	89%
42	Claude Sonnet 4.5	$0.045	41.1s	82%	94	90	88	87	83	88%
19	DeepSeek V4 Flash (Reasoning)	$0.0007	25.0s	83%	93	90	88	86	86	88%
13	Muse Spark 1.1 (Reasoning, Medium)	$0.016	22.4s	86%	91	90	89	86	86	88%
37	GPT-5.6 Sol	$0.093	56.9s	88%	89	89	88	88	87	88%
14	GPT-5.4 Mini (Reasoning, Low)	$0.014	16.1s	85%	91	90	88	88	85	88%
62	Claude Sonnet 4	$0.045	54.2s	79%	96	91	90	86	78	88%
109	Grok 4.7 (Reasoning, High)	$0.090	4.2m	87%	89	89	88	87	87	88%
64	GPT-5.5 (Reasoning, Low)	$0.122	1.5m	87%	89	89	89	89	86	88%
172	Qwen 3.8 Max (Reasoning, XHigh)	$0.098	6.8m	85%	90	89	88	87	86	88%
43	Z.AI GLM 5	$0.012	1.8m	83%	92	90	88	86	83	88%
50	Kimi K3 (Reasoning, High)	$0.043	1.7m	83%	92	90	89	85	84	88%
22	GPT-6 Sol	$0.025	43.5s	86%	89	89	88	87	86	88%
68	DeepSeek V4 Flash 0731 (Reasoning, High)	$0.0067	2.9m	82%	93	89	88	87	83	88%
27	Claude Haiku 5.5 (Adaptive)	$0.0024	21.8s	82%	92	91	87	87	83	88%
55	Claude Opus 4.7	$0.085	30.7s	82%	93	90	89	85	82	88%
195	Claude Opus 4	$0.434	2.8m	80%	94	89	86	85	85	88%
25	Claude Haiku 5.5 (Reasoning, Medium)	$0.0017	14.1s	82%	92	90	88	86	83	88%
105	GPT-5.4 (Reasoning)	$0.092	2.6m	83%	91	89	88	87	84	88%
41	Z.AI GLM 5.3 Flash (Reasoning, Low)	$0.0006	1.0m	81%	94	91	91	85	78	88%
28	Muse Spark 1.2 (Reasoning, Medium)	$0.018	37.3s	83%	91	89	87	87	84	88%
69	Aion 3.0	$0.031	1.1m	78%	95	90	87	84	82	88%
48	Claude Opus 5 (Reasoning, Low)	$0.103	54.1s	86%	89	88	88	87	86	88%
17	DeepSeek V4 Pro	$0.0027	1.1m	87%	88	88	88	87	87	88%
67	Muse Spark 1.3 (Reasoning, Medium)	$0.028	2.1m	82%	91	89	86	86	86	88%
59	Z.AI GLM 5.1	$0.014	1.2m	80%	93	91	87	85	81	87%
35	Xiaomi MIMO v2.5 Pro	$0.013	1.2m	84%	90	88	87	86	86	87%
46	Kimi K3 (Reasoning, Low)	$0.029	1.2m	82%	91	90	87	85	83	87%
80	Aion 3.5 Mini (Reasoning, High)	$0.014	2.9m	83%	91	90	88	85	82	87%
21	GPT-6 Luna	$0.0012	24.4s	84%	90	88	87	86	85	87%
36	Grok 4.20 (Reasoning)	$0.014	1.0m	83%	91	89	89	85	82	87%
33	Qwen 3 235B A22B Instruct 2507	$0.0011	49.4s	83%	90	89	86	85	85	87%
34	Grok 4.20	$0.011	46.5s	83%	90	89	89	87	81	87%
127	Qwen 3.5 397B A17B	$0.0045	6.2m	85%	88	88	87	87	86	87%
40	Muse Spark 1.1 (Reasoning, Minimal)	$0.014	18.4s	80%	93	88	86	84	84	87%
31	GPT-6 Luna (Reasoning, High)	$0.0023	58.8s	84%	89	88	87	85	84	87%
45	Xiaomi MIMO v2.6 Flash	$0.0009	1.9m	84%	89	88	87	85	84	87%
26	GPT-5.4 Mini (Reasoning)	$0.022	26.8s	85%	88	88	86	86	86	87%
38	Qwen 3.6 Flash	$0.013	52.6s	83%	90	87	87	86	84	87%
23	Gemini 3.5 Flash Lite (Reasoning, Minimal)	$0.0041	6.1s	84%	89	88	87	86	84	87%
98	Claude Sonnet 4.6 (Reasoning)	$0.085	1.3m	82%	89	89	88	88	80	87%
60	Claude Opus 4.8 (Reasoning)	$0.083	39.5s	83%	89	89	88	84	82	87%
95	Claude Opus 4.5	$0.086	54.6s	80%	93	86	86	85	84	87%
116	Qwen 3.6 Max Preview	$0.057	3.7m	84%	89	88	87	86	84	87%
79	Claude Opus 4.6	$0.088	1.2m	83%	89	88	87	85	84	87%
73	Grok 4.5 (Reasoning, High)	$0.035	1.9m	82%	89	89	86	86	83	86%
44	DeepSeek V4 Flash	$0.0005	17.3s	79%	93	89	86	82	82	86%
137	Kimi K2.5	$0.019	5.2m	82%	90	89	88	83	81	86%
128	Qwen 3.8 27B (Reasoning, XHigh)	$0.043	4.5m	83%	88	87	86	85	85	86%
76	Qwen 3.5 Plus (2026-04-20)	$0.018	1.7m	81%	91	86	85	85	84	86%
136	Hy4 Preview (Reasoning, High)	$0.046	4.7m	83%	89	88	87	83	82	86%
88	Claude Opus 4.8 (Reasoning, Low)	$0.084	40.7s	81%	90	87	86	85	82	86%
70	DeepSeek V4 Pro 0813 (Reasoning, High)	$0.021	1.3m	81%	91	88	86	83	82	86%
75	Aion 3.0 Mini	$0.0071	1.3m	79%	91	89	85	84	82	86%
97	GPT-5.1	$0.054	1.7m	82%	90	87	86	84	82	86%
103	Claude Sonnet 5 (Reasoning)	$0.038	42.7s	76%	94	86	84	83	82	86%
66	Claude Opus 4.7 (Reasoning)	$0.086	28.8s	83%	88	86	85	85	84	86%
85	DeepSeek V4 Pro (Reasoning)	$0.016	2.1m	82%	89	87	86	85	82	86%
72	Claude Sonnet 5 (Reasoning, Low)	$0.040	43.3s	80%	89	87	84	84	83	86%
119	Qwen 3.8 Flash (Reasoning)	$0.0079	3.4m	80%	90	87	86	84	81	86%
47	Mistral Medium 3.5	$0.025	24.4s	82%	89	87	85	84	83	86%
39	Muse Glimmer 30B (Reasoning, Medium)	$0.0042	47.0s	83%	88	86	86	85	83	86%
104	Claude Opus 5	$0.093	47.0s	81%	89	88	87	84	79	85%
83	MiniMax M2.5	$0.0043	1.8m	81%	89	87	86	85	80	85%
134	MiniMax M3	$0.0089	3.8m	79%	89	88	84	82	82	85%
89	Z.AI GLM 5 Turbo	$0.013	40.7s	77%	92	88	85	81	79	85%
51	Grok 4.3	$0.0088	25.8s	80%	88	87	84	83	82	85%
52	Qwen 3.5 Flash	$0.0024	35.9s	80%	89	86	86	84	79	85%
86	Grok 4.3 (Reasoning)	$0.016	1.7m	81%	88	85	85	84	82	85%
78	Laguna S 2.1	$0.0005	34.9s	77%	90	89	84	80	80	85%
82	MiniMax M2.7	$0.0053	1.3m	80%	88	87	84	82	81	84%
87	Qwen 3.6 35B	$0.0075	1.0m	79%	89	86	84	83	80	84%
81	Claude Sonnet 4.6	$0.038	37.6s	80%	86	86	84	84	80	84%
101	Mistral Medium 3.5 (Reasoning)	$0.024	22.4s	77%	90	87	83	81	80	84%
196	Kimi K2.6	$0.085	9.1m	79%	87	86	83	82	81	84%
84	Hermes 3 405B	$0.0054	49.2s	79%	88	87	87	81	76	84%
133	GPT-5	$0.059	2.8m	80%	87	84	83	83	82	84%
77	Gemini 3.5 Flash Lite (Reasoning)	$0.0040	5.6s	77%	89	87	84	82	78	84%
149	Gemini 3.1 Pro (Preview)	$0.134	2.1m	81%	87	84	84	83	81	84%
92	Mistral Large 4.0	$0.0083	1.3m	80%	88	85	84	82	80	84%
61	Mistral Medium 3.1	$0.0058	42.8s	81%	86	83	83	83	83	84%
91	Qwen 3.5 27B	$0.012	58.0s	79%	86	85	83	82	81	83%
65	Gemini 3.7 Flash (Reasoning, Medium)	$0.014	21.4s	80%	85	85	85	84	78	83%
54	GPT-5.4 Nano (Reasoning)	$0.0055	23.1s	81%	86	85	85	81	80	83%
102	o4 Mini	$0.015	26.6s	77%	90	83	82	81	81	83%
58	GPT-5.4 Nano	$0.0050	19.6s	81%	85	84	83	82	82	83%
63	Mistral Small 4 (Reasoning)	$0.0027	32.6s	81%	85	84	83	83	81	83%
125	WizardLM 2 8x22b	$0.0042	3.0m	81%	85	84	84	81	81	83%
140	Qwen 3.7 Max	$0.078	2.3m	80%	85	83	82	82	82	83%
96	Hermes 3 70B	$0.0015	38.9s	78%	87	85	84	83	76	83%
115	Thinking Machines Inkling	$0.0093	1.9m	79%	86	84	83	83	79	83%
189	Thinking Machines Inkling (Reasoning)	$0.028	6.6m	78%	88	84	84	80	78	83%
99	GPT-5.4 Nano (Reasoning, Low)	$0.0045	16.6s	77%	87	86	82	81	78	83%
120	DeepSeek V3.2	$0.0019	1.3m	76%	88	85	82	80	79	83%
122	Qwen 3.5 Plus (2026-02-15)	$0.0071	31.1s	74%	91	86	83	77	76	83%
144	ByteDance Seed 1.6	$0.010	1.8m	73%	89	88	81	79	76	83%
93	Qwen 3.5 122B	$0.020	45.8s	80%	84	83	82	82	81	82%
157	Mistral Large 4.0 (Reasoning)	$0.027	3.6m	79%	85	83	82	81	80	82%
110	Gemma 4 31B (Reasoning)	$0.0016	1.2m	78%	86	82	81	81	80	82%
100	DeepSeek V3 (2025-03-24)	$0.0017	44.1s	79%	85	83	81	81	80	82%
153	Cohere Command R+ (Aug. 2024)	$0.028	52.8s	71%	89	85	80	80	72	82%
111	Z.AI GLM 4.5 Air	$0.0024	36.5s	76%	87	82	81	79	79	82%
112	Mistral Large 2	$0.017	30.9s	77%	86	83	82	79	77	81%
108	Gemini 3.6 Flash (Reasoning)	$0.042	28.3s	80%	83	83	82	80	79	81%
94	Gemini 3.8 Flash (Reasoning, Medium)	$0.014	16.2s	79%	83	83	81	80	79	81%
118	Xiaomi MIMO v2.5	$0.0055	29.8s	76%	85	84	81	81	76	81%
160	Qwen 3.6 27B	$0.036	3.0m	78%	85	82	82	79	78	81%
129	o4 Mini High	$0.026	48.1s	75%	86	81	80	79	79	81%
107	Mistral Large 3	$0.0048	36.1s	78%	84	83	81	80	78	81%
90	Ministral 3 14B	$0.0012	11.9s	79%	83	82	82	80	79	81%
114	Qwen 3 32B	$0.0013	32.4s	76%	85	83	82	80	74	81%
113	ByteDance Seed 1.6 Flash	$0.0015	30.8s	77%	85	82	82	79	76	81%
106	Mistral Small 4	$0.0013	11.7s	77%	84	81	81	81	76	81%
124	Gemini 2.5 Flash	$0.0065	12.1s	74%	85	85	81	77	75	81%
130	Aion 2.0	$0.0079	1.2m	76%	85	83	82	78	76	81%
117	Qwen 3.5 9B	$0.0011	1.1m	78%	83	82	82	79	77	81%
126	GPT-5 Mini	$0.0098	1.1m	77%	84	81	80	79	78	81%
121	Qwen 3.5 35B	$0.018	1.2m	79%	82	82	81	79	79	81%
123	Z.AI GLM 4.7 Flash	$0.0018	1.1m	77%	83	83	83	79	75	80%
170	ByteDance Seed 2.0 Mini	$0.0044	4.4m	80%	81	81	80	80	80	80%
135	Z.AI GLM 4.7	$0.012	1.7m	77%	83	81	80	79	78	80%
131	Gemini 3 Flash (Preview, Reasoning)	$0.012	35.6s	74%	85	82	79	79	76	80%
148	Gemini 3.5 Flash (Reasoning, Minimal)	$0.019	12.1s	71%	87	82	77	77	76	80%
145	DeepSeek V3 (2024-12-26)	$0.0027	1.4m	74%	85	80	79	79	75	80%
159	Claude Haiku 4.5	$0.015	24.9s	70%	89	78	78	77	76	79%
142	Claude Sonnet 5	$0.034	35.0s	75%	85	81	80	77	75	79%
168	GPT-5.2	$0.056	1.5m	75%	83	80	78	78	77	79%
162	Laguna XS 2.1	$0.0005	17.0s	69%	86	84	77	77	71	79%
138	GPT-4.1	$0.020	38.4s	75%	81	80	79	79	74	79%
132	Gemini 2.5 Pro	$0.037	35.2s	77%	80	80	79	78	77	79%
155	Gemma 4 31B	$0.0014	1.3m	74%	84	79	79	77	74	79%
177	DeepSeek V3.1	$0.0024	2.3m	73%	83	80	79	78	73	79%
151	Z.AI GLM 4.5	$0.0061	35.2s	72%	83	81	78	78	72	78%
139	Gemini 2.5 Flash (Reasoning)	$0.011	20.8s	74%	82	80	78	77	75	78%
141	Gemini 3.6 Flash (Reasoning, Minimal)	$0.019	13.6s	74%	81	80	78	77	74	78%
147	Z.AI GLM 4.6	$0.0097	46.5s	74%	82	80	78	76	75	78%
176	Cydonia 24B V4.1	$0.0023	1.1m	70%	85	81	79	77	68	78%
178	ByteDance Seed 2.0 Lite	$0.011	1.7m	72%	84	80	77	75	74	78%
152	GPT-4o, Aug. 6th (temp=1)	$0.020	16.0s	73%	83	80	78	76	73	78%
174	Gemini 3.5 Flash (Reasoning)	$0.086	44.8s	75%	80	80	79	76	75	78%
158	Ministral 8B	$0.0007	20.3s	71%	86	79	78	76	71	78%
175	Gemma 4 26B	$0.0012	43.6s	69%	84	82	81	78	64	78%
146	Mistral NeMO	$0.0008	7.5s	72%	82	82	80	73	70	77%
165	Gemma 3 27B	$0.0010	57.7s	72%	82	81	78	73	73	77%
161	Gemma 3 12B	$0.0004	45.9s	73%	81	79	77	75	74	77%
154	Gemma 4 26B (Reasoning)	$0.0019	49.8s	74%	80	79	78	75	74	77%
173	Gemini 3.1 Flash Lite (Preview)	$0.0037	8.7s	69%	85	79	78	77	67	77%
156	Ministral 3 3B	$0.0006	6.6s	72%	81	79	78	76	70	77%
143	Gemini 3.1 Flash Lite	$0.0036	9.3s	74%	79	77	76	76	75	76%
150	Qwen 2.5 72B	$0.0012	30.2s	74%	78	77	76	75	74	76%
164	GPT-4.1 Mini	$0.0034	20.9s	72%	80	79	76	74	73	76%
182	DeepSeek-V2 Chat	$0.0026	52.7s	68%	85	76	74	73	73	76%
166	Arcee AI: Trinity Mini	$0.0004	9.1s	71%	81	76	75	75	72	76%
171	Gemini 3 Flash (Preview)	$0.0089	20.4s	72%	79	78	76	73	72	76%
185	Thinking Machines Inkling Small (Reasoning, High)	$0.012	1.8m	71%	80	77	76	72	72	75%
167	GPT-4o, Aug. 6th (temp=0)	$0.021	12.5s	73%	78	76	75	74	74	75%
169	Ministral 3B	$0.0002	7.0s	72%	77	76	75	74	71	75%
188	Ministral 3 8B	$0.0012	22.3s	66%	80	78	72	71	69	74%
179	Gemini 2.5 Flash Lite	$0.0012	10.8s	70%	78	76	74	73	70	74%
181	Inception Mercury 2	$0.0029	6.1s	70%	76	76	73	72	72	74%
183	GPT-4o Mini (temp=0)	$0.0016	32.9s	69%	77	73	72	72	72	73%
184	GPT-4o Mini (temp=1)	$0.0014	39.7s	70%	77	75	74	72	69	73%
187	Llama 3.1 70B	$0.0025	16.4s	68%	77	76	76	68	65	72%
193	GPT-5 Nano	$0.0042	1.5m	66%	78	73	71	69	67	72%
186	Gemma 3 4B	$0.0003	21.4s	70%	74	73	72	70	70	72%
190	Gemini 3.1 Flash Lite (Reasoning)	$0.0034	10.1s	68%	73	73	72	72	65	71%
192	GPT-OSS 120B	$0.0010	1.6m	68%	72	71	70	70	68	70%
194	Nemotron 3 Super	$0.0000	40.4s	66%	72	71	68	68	66	69%
191	GPT-4.1 Nano	$0.0010	14.3s	66%	74	70	70	67	65	69%
199	Mistral Small 3.2 24B	$0.0078	6.5m	59%	81	68	67	67	63	69%
198	Gemini 2.5 Flash Lite (Reasoning)	$0.0036	40.3s	61%	78	71	69	64	62	69%
197	Nemotron 3 Nano	$0.0016	1.7m	64%	73	73	71	65	61	69%
83.27%

Median	Evaluator	Top 3	Flop 3
100.0%	"Not X but Y" pattern overuse	100Claude Sonnet 5 (Reasoning) 100Mistral NeMO 100GPT-5.6 Terra (Reasoning, Medium)	13Thinking Machines Inkling Small (Reasoning, High) 45GPT-5 Nano 50Aion 2.0
56.0%	Adverb-first sentence starts	100Claude Haiku 5.5 (Reasoning, Low) 100GPT-5.4 Nano (Reasoning, Low) 98Claude Sonnet 5.5 (Reasoning)	0Llama 3.1 70B 0Inception Mercury 2 0Gemini 3.1 Flash Lite (Reasoning)
100.0%	Adverbs in dialogue tags	100ByteDance Seed 1.6 Flash 100GPT-6 Luna 100Qwen 3.5 Plus (2026-04-20)	19GPT-4.1 Nano 26Cydonia 24B V4.1 48Llama 3.1 70B
90.3%	AI-ism adverb frequency	100Grok 4.7 (Reasoning, High) 99GPT-6 Sol 99o4 Mini	52Gemma 3 4B 54Cydonia 24B V4.1 67GPT-4.1 Nano
100.0%	AI-ism character names	100DeepSeek V4 Flash (Reasoning) 100Z.AI GLM 5.3 Flash (Reasoning, Low) 100Qwen 3 235B A22B Instruct 2507	60Claude Opus 4 88Z.AI GLM 5 92Aion 3.0 Mini
100.0%	AI-ism location names	100Qwen 3.5 35B 100DeepSeek V4.1 Flash (Reasoning, High) 100Cohere Command R+ (Aug. 2024)	96Gemini 2.5 Pro 96Z.AI GLM 4.5 Air
59.7%	AI-ism word frequency	92GPT-6 Sol (Reasoning, Medium) 91Claude Sonnet 4.6 (Reasoning) 91Claude Opus 5 (Reasoning)	0Gemma 3 4B 0GPT-4o Mini (temp=0) 0Inception Mercury 2
100.0%	Cliché density	100Grok 4.7 (Reasoning, High) 100Gemini 3 Flash (Preview) 100GPT-5	13GPT-4o Mini (temp=0) 13Mistral Small 3.2 24B 47Gemini 2.5 Flash Lite
99.1%	Dialogue tag variety (said vs. fancy)	100GPT-4.1 100Z.AI GLM 5.3 Flash (Reasoning, Max) 100DeepSeek V4 Pro (Reasoning)	11Gemini 2.5 Flash Lite 18GPT-OSS 120B 24GPT-4o, Aug. 6th (temp=1)
100.0%	Em-dash & semicolon overuse	100Qwen 3.6 Flash 100Claude Opus 4.7 100Grok 4.7 (Reasoning, High)	0GPT-4o, Aug. 6th (temp=1) 0Mistral Small 4 0GPT-5 Nano
100.0%	Emotion telling (show vs. tell)	100Gemini 2.5 Pro 100Claude Sonnet 5 (Reasoning) 100Z.AI GLM 5	87Mistral Small 3.2 24B 88GPT-4o Mini (temp=0) 97Qwen 2.5 72B
100.0%	Filter word density	100Claude Opus 5 (Reasoning) 100DeepSeek V3 (2024-12-26) 100GPT-5.6 Sol (Reasoning, Medium)	34ByteDance Seed 2.0 Mini 52Nemotron 3 Nano 62Gemini 3.1 Flash Lite (Reasoning)
100.0%	Gibberish response detection	100Qwen 3.5 27B 100Claude Opus 4.8 (Reasoning) 100GPT-5.2	80Llama 3.1 70B 91Cydonia 24B V4.1 99DeepSeek V3 (2025-03-24)
100.0%	Markdown formatting overuse	100Claude Haiku 5.5 (Adaptive) 100DeepSeek V3.2 100GPT-5.5 (Reasoning, Medium)	80Llama 3.1 70B 85Qwen 3 32B 93Cydonia 24B V4.1
100.0%	Missing dialogue indicators (quotation marks)	100GPT-4o Mini (temp=0) 100Mistral Large 3 100Claude Opus 4.7	63Qwen 3.5 Flash 80Qwen 3.5 Plus (2026-02-15) 80Qwen 3.7 Flash (Reasoning)
86.5%	Name drop frequency	100Claude Opus 5.5 (Reasoning) 100Aion 3.5 Mini (Reasoning, High) 100Gemini 2.5 Flash	9GPT-5.2 9Qwen 3.5 9B 15GPT-6 Sol (Reasoning, Medium)
90.4%	Narrator intent-glossing	100Claude Opus 5.5 (Reasoning) 100Muse Spark 1.2 (Reasoning, Medium) 100Qwen 3.6 Flash	22Nemotron 3 Super 25Claude Sonnet 5 26Claude Haiku 4.5
100.0%	Overuse of "that" (subordinate clause padding)	100Claude Opus 4.7 (Reasoning) 100Claude Opus 4 100Z.AI GLM 4.7	49Mistral Small 3.2 24B 64ByteDance Seed 2.0 Mini 76Claude Haiku 5.5 (Adaptive)
100.0%	Paragraph length variance	100Gemini 2.5 Flash (Reasoning) 100GPT-5.1 100GPT-5.6 Terra (Reasoning, Medium)	64Nemotron 3 Nano 67GPT-OSS 120B 67Inception Mercury 2
99.7%	Passive voice overuse	100Qwen 3.6 35B 100Qwen 3.5 Plus (2026-04-20) 100o4 Mini High	80ByteDance Seed 2.0 Lite 88Qwen 3.5 35B 93Hermes 3 70B
100.0%	Past progressive (was/were + -ing) overuse	100GPT-5.6 Sol (Reasoning, Medium) 100Grok 4.3 (Reasoning) 100Gemini 3.1 Pro (Preview)	57Claude Opus 5 60Llama 3.1 70B 60Z.AI GLM 4.6
50.5%	Pronoun-first sentence starts	100GPT-5.5 (Reasoning, Low) 100GPT-5.6 Sol (Reasoning, Medium) 100GPT-5.6 Terra (Reasoning, Medium)	0Gemini 3 Flash (Preview, Reasoning) 0Gemini 3.1 Flash Lite (Preview) 0ByteDance Seed 2.0 Mini
97.6%	Purple prose (modifier overload)	100Qwen 2.5 72B 100Inception Mercury 2 100Qwen 3.8 Max (Reasoning, XHigh)	58Cydonia 24B V4.1 69Gemma 3 4B 84Gemini 3.5 Flash (Reasoning)
100.0%	Repeated phrase echo	100Claude Haiku 5.5 (Adaptive) 100Claude Sonnet 4.6 (Reasoning) 100Claude Sonnet 5 (Reasoning, Low)	—
100.0%	Sentence length variance	100Claude Opus 4.8 (Reasoning) 100Claude Opus 4.7 100Muse Glimmer 30B (Reasoning, Medium)	74Mistral Small 3.2 24B 92Nemotron 3 Nano 94Qwen 3.5 35B
56.5%	Sentence opener variety	97Cydonia 24B V4.1 92GPT-4o, Aug. 6th (temp=1) 91Qwen 3.7 Flash (Reasoning)	32Qwen 3.5 35B 37Gemini 3 Flash (Preview, Reasoning) 37Qwen 3.5 Flash
22.3%	Subject-first sentence starts	92Qwen 3.7 Flash (Reasoning) 89Cydonia 24B V4.1 85Claude Haiku 5.5 (Reasoning, Low)	0Gemma 4 26B 0Gemma 4 31B (Reasoning) 0Inception Mercury 2
20.0%	Subordinate conjunction sentence starts	100Claude Haiku 5.5 (Reasoning, Low) 90Cohere Command R+ (Aug. 2024) 80Aion 3.5 (Reasoning, High)	0Gemini 3.5 Flash Lite (Reasoning, Minimal) 0Ministral 3 8B 0GPT-4.1
86.2%	Technical jargon density	100Gemini 3.1 Pro (Preview) 100GPT-6 Sol 100o4 Mini High	0GPT-4o Mini (temp=1) 2Ministral 3B 7ByteDance Seed 2.0 Lite
78.8%	Useless dialogue additions	100Gemini 3.1 Pro (Preview) 100Qwen 3.5 27B 100Qwen 3.8 27B (Reasoning, XHigh)	0Ministral 8B 0Ministral 3 3B 0Inception Mercury 2

Bad Writing Habits

Romance: separated couple reunites

Performance Score Distribution (Top 20)

Price-Performance Score Distribution (Top 20)

Most Stable Models (Top 20)

Top Overall Models (Top 20)