10 codex entries

Codex Violation Detection

Detects factual inconsistencies between a story bible and prose passages. The model must output structured XML identifying each violation with paragraph number and substring.

Tooling Reasoning

Performance Score Distribution (Top 20)

Click a model name to view its detail page.

	Score
Claude Opus 4.6 (Reasoning)	100%
Gemini 3.1 Pro (Preview)	100%
Z.AI GLM 5 Turbo	100%
Claude Opus 4.6	100%
Qwen 3.5 397B A17B	100%
GPT-5.4 (Reasoning, Low)	100%
MoonshotAI: Kimi K2.5	100%
Qwen 3.5 27B	100%
Gemini 3 Flash (Preview, Reasoning)	100%
Claude Opus 4.5	100%
Grok 4.1 Fast	100%
Gemini 3 Pro (Preview)	100%
Claude Sonnet 4	100%
Z.AI GLM 4.7	100%
Claude Sonnet 4.5	100%
Qwen 3.5 35B	100%
ByteDance Seed 2.0 Mini	100%
Gemini 2.5 Flash (Reasoning)	100%
Gemini 3 Flash (Preview)	100%
ByteDance Seed 2.0 Lite	100%

	Score	Cost	Time
Stealth: Aurora Alpha	100%	—	4.5s
Gemini 3.1 Flash Lite (Preview)	94%	$0.0011	1.5s
Gemini 3 Flash (Preview)	100%	$0.0015	2.7s
Mistral Medium 3.1	99%	$0.0015	3.7s
Grok 4 Fast	97%	$0.0010	7.8s
Inception Mercury 2	98%	$0.0020	3.2s
Stealth: Healer Alpha	97%	$0.0000	19.0s
Qwen3 235B A22B Instruct 2507	96%	$0.0004	10.3s
Qwen 3.5 Plus (2026-02-15)	94%	$0.0018	12.4s
Mistral Small Creative	93%	$0.0003	2.3s
Grok 4.1 Fast	100%	$0.0013	18.5s
Mistral Small 4 (Reasoning)	93%	$0.0014	11.4s
MiniMax M2.5	97%	$0.0014	17.9s
Z.AI GLM 5 Turbo	100%	$0.0035	9.0s
Writer: Palmyra X5	94%	$0.0034	6.6s
GPT-5.4 Nano (Reasoning)	87%	$0.0018	7.4s
ByteDance Seed 1.6 Flash	94%	$0.0005	6.5s
MiniMax M2.7	98%	$0.0014	19.7s
GPT-5.4	100%	$0.0061	4.2s
Gemini 2.5 Flash Lite (Reasoning)	95%	$0.0014	10.7s

	Score	Cost	Speed	Stability
Gemini 3 Flash (Preview)	100%	$0.0015	2.7s	100%
Stealth: Aurora Alpha	100%	—	4.5s	100%
Mistral Medium 3.1	99%	$0.0015	3.7s	96%
Z.AI GLM 5 Turbo	100%	$0.0035	9.0s	100%
GPT-5.4	100%	$0.0061	4.2s	100%
Inception Mercury 2	98%	$0.0020	3.2s	94%
Gemini 2.5 Flash (Reasoning)	100%	$0.0049	7.5s	100%
Grok 4.1 Fast	100%	$0.0013	18.5s	100%
Gemini 3 Flash (Preview, Reasoning)	100%	$0.0067	11.1s	100%
GPT-5.4 (Reasoning, Low)	100%	$0.0096	6.3s	100%
Qwen3 235B A22B Instruct 2507	96%	$0.0004	10.3s	90%
ByteDance Seed 1.6 Flash	94%	$0.0005	6.5s	90%
Claude Sonnet 4	100%	$0.012	4.6s	100%
Claude Sonnet 4.5	100%	$0.012	5.1s	100%
Llama 3.1 Nemotron 70B	97%	$0.0030	9.7s	91%
Mistral Small Creative	93%	$0.0003	2.3s	83%
Gemini 2.5 Flash Lite (Reasoning)	95%	$0.0014	10.7s	89%
GPT-4o, Aug. 6th (temp=0)	94%	$0.0063	2.2s	91%
GPT-5.4 Mini (Reasoning, Low)	95%	$0.0029	4.0s	86%
MiniMax M2.7	98%	$0.0014	19.7s	93%

Rank	Model	Avg. Cost	Avg. Time	Stability	# 1	# 2	# 3	# 4	# 5	# 6	# 7	# 8	# 9	# 10	Total
79	Claude Opus 4.6 (Reasoning)	$0.034	15.8s	100%	100	100	100	100	100	100	100	100	100	100	100%
81	Gemini 3.1 Pro (Preview)	$0.031	24.2s	100%	100	100	100	100	100	100	100	100	100	100	100%
4	Z.AI GLM 5 Turbo	$0.0035	9.0s	100%	100	100	100	100	100	100	100	100	100	100	100%
43	Claude Opus 4.6	$0.021	6.0s	100%	100	100	100	100	100	100	100	100	100	100	100%
109	Qwen 3.5 397B A17B	$0.018	2.0m	100%	100	100	100	100	100	100	100	100	100	100	100%
10	GPT-5.4 (Reasoning, Low)	$0.0096	6.3s	100%	100	100	100	100	100	100	100	100	100	100	100%
62	MoonshotAI: Kimi K2.5	$0.0093	48.0s	100%	100	100	100	100	100	100	100	100	100	100	100%
86	Qwen 3.5 27B	$0.014	1.1m	100%	100	100	100	100	100	100	100	100	100	100	100%
9	Gemini 3 Flash (Preview, Reasoning)	$0.0067	11.1s	100%	100	100	100	100	100	100	100	100	100	100	100%
42	Claude Opus 4.5	$0.021	5.7s	100%	100	100	100	100	100	100	100	100	100	100	100%
8	Grok 4.1 Fast	$0.0013	18.5s	100%	100	100	100	100	100	100	100	100	100	100	100%
80	Gemini 3 Pro (Preview)	$0.032	21.4s	100%	100	100	100	100	100	100	100	100	100	100	100%
13	Claude Sonnet 4	$0.012	4.6s	100%	100	100	100	100	100	100	100	100	100	100	100%
21	Z.AI GLM 4.7	$0.0047	26.8s	100%	100	100	100	100	100	100	100	100	100	100	100%
14	Claude Sonnet 4.5	$0.012	5.1s	100%	100	100	100	100	100	100	100	100	100	100	100%
76	Qwen 3.5 35B	$0.014	51.2s	100%	100	100	100	100	100	100	100	100	100	100	100%
96	ByteDance Seed 2.0 Mini	$0.0020	1.8m	100%	100	100	100	100	100	100	100	100	100	100	100%
7	Gemini 2.5 Flash (Reasoning)	$0.0049	7.5s	100%	100	100	100	100	100	100	100	100	100	100	100%
1	Gemini 3 Flash (Preview)	$0.0015	2.7s	100%	100	100	100	100	100	100	100	100	100	100	100%
44	ByteDance Seed 2.0 Lite	$0.0044	45.7s	100%	100	100	100	100	100	100	100	100	100	100	100%
5	GPT-5.4	$0.0061	4.2s	100%	100	100	100	100	100	100	100	100	100	100	100%
2	Stealth: Aurora Alpha	—	4.5s	100%	100	100	100	100	100	100	100	100	100	100	100%
35	ByteDance Seed 1.6	$0.0034	31.9s	95%	100	100	100	100	100	100	100	100	100	92	99%
34	GPT-5.2	$0.013	12.3s	95%	100	100	100	100	100	100	100	100	100	92	99%
59	Aion 2.0	$0.0052	48.7s	95%	100	100	100	100	100	100	100	100	100	92	99%
61	Gemini 2.5 Pro	$0.022	14.8s	95%	100	100	100	100	100	100	100	100	100	92	99%
94	Grok 4	$0.030	38.7s	95%	100	100	100	100	100	100	100	100	100	92	99%
104	Claude Opus 4	$0.062	8.6s	95%	100	100	100	100	100	100	100	100	100	92	99%
56	Claude 3.5 Sonnet	$0.024	6.9s	96%	100	100	100	100	100	100	100	100	94	94	99%
3	Mistral Medium 3.1	$0.0015	3.7s	96%	100	100	100	100	100	100	100	100	94	94	99%
103	Qwen 3.5 9B	$0.0015	2.1m	95%	100	100	100	100	100	100	100	100	94	92	99%
69	Z.AI GLM 5	$0.0079	41.9s	92%	100	100	100	100	100	100	100	100	100	86	99%
99	Qwen 3.5 122B	$0.021	1.1m	94%	100	100	100	100	100	100	100	100	92	92	98%
54	Grok 4.20 (Beta, Reasoning)	$0.020	11.6s	94%	100	100	100	100	100	100	100	100	92	92	98%
6	Inception Mercury 2	$0.0020	3.2s	94%	100	100	100	100	100	100	100	100	92	92	98%
30	GPT-4o, May 13th (temp=0)	$0.015	2.2s	95%	100	100	100	100	100	100	100	94	94	94	98%
20	MiniMax M2.7	$0.0014	19.7s	93%	100	100	100	100	100	100	100	94	94	89	98%
78	GPT-5.4 (Reasoning)	$0.021	30.3s	93%	100	100	100	100	100	100	100	92	92	92	98%
28	GPT-5.4 Mini (Reasoning)	$0.0082	10.2s	93%	100	100	100	100	100	100	94	94	94	89	97%
22	Stealth: Healer Alpha	$0.0000	19.0s	89%	100	100	100	100	100	100	100	100	86	86	97%
15	Llama 3.1 Nemotron 70B	$0.0030	9.7s	91%	100	100	100	100	100	94	94	94	94	92	97%
24	Grok 4 Fast	$0.0010	7.8s	82%	100	100	100	100	100	100	100	100	100	69	97%
27	MiniMax M2.5	$0.0014	17.9s	89%	100	100	100	100	100	100	100	94	86	86	97%
11	Qwen3 235B A22B Instruct 2507	$0.0004	10.3s	90%	100	100	100	100	100	100	92	92	92	86	96%
66	o4 Mini	$0.014	20.7s	89%	100	100	100	100	100	100	92	92	86	86	96%
85	GPT-5.1	$0.020	29.2s	86%	100	100	100	100	92	92	92	92	92	92	95%
111	GPT-5	$0.042	1.0m	86%	100	100	100	100	92	92	92	92	92	92	95%
83	Qwen 3.5 Flash	$0.0031	55.2s	82%	100	100	100	100	100	100	100	92	86	72	95%
19	GPT-5.4 Mini (Reasoning, Low)	$0.0029	4.0s	86%	100	100	100	100	94	94	94	89	89	89	95%
97	o4 Mini High	$0.024	36.5s	86%	100	100	100	100	100	92	92	92	86	86	95%
17	Gemini 2.5 Flash Lite (Reasoning)	$0.0014	10.7s	89%	100	100	94	94	94	94	94	92	92	92	95%
29	Qwen 3.5 Plus (2026-02-15)	$0.0018	12.4s	86%	100	100	100	100	100	100	86	86	86	86	94%
26	Writer: Palmyra X5	$0.0034	6.6s	86%	100	100	100	100	100	100	86	86	86	86	94%
12	ByteDance Seed 1.6 Flash	$0.0005	6.5s	90%	100	94	94	94	94	94	94	94	92	89	94%
23	GPT-4o, Aug. 6th (temp=1)	$0.0062	2.3s	89%	100	94	94	94	94	94	94	94	89	89	94%
18	GPT-4o, Aug. 6th (temp=0)	$0.0063	2.2s	91%	94	94	94	94	94	94	94	94	94	89	94%
25	Gemini 3.1 Flash Lite (Preview)	$0.0011	1.5s	80%	100	100	100	100	100	100	100	86	79	72	94%
32	Z.AI GLM 4.5	$0.0014	10.9s	84%	100	100	100	100	100	92	86	86	86	86	94%
37	DeepSeek V3 (2025-03-24)	$0.0008	16.3s	84%	100	100	100	94	94	94	92	86	86	86	93%
16	Mistral Small Creative	$0.0003	2.3s	83%	100	100	100	94	94	94	92	92	86	79	93%
38	Qwen 3 32B	$0.0006	11.1s	80%	100	100	100	94	94	94	94	94	89	72	93%
71	GPT-5 Mini	$0.0072	38.4s	88%	100	92	92	92	92	92	92	92	92	92	93%
95	Claude Sonnet 4.6 (Reasoning)	$0.027	19.3s	80%	100	100	100	100	100	86	86	86	86	86	93%
40	Mistral Small 4 (Reasoning)	$0.0014	11.4s	80%	100	100	100	100	100	100	92	86	79	72	93%
67	Stealth: Hunter Alpha	$0.0000	32.8s	78%	100	100	100	100	94	86	86	86	86	86	92%
84	Z.AI GLM 4.6	$0.0025	40.3s	72%	100	100	100	100	100	100	92	92	86	53	92%
41	Inception Mercury	$0.0003	7.9s	77%	100	100	100	100	94	86	86	86	86	79	92%
101	GPT-5 Nano	$0.0037	1.4m	76%	100	100	94	86	86	86	86	86	86	86	90%
72	Z.AI GLM 4.7 Flash	$0.0010	36.3s	80%	100	94	89	89	89	89	89	89	89	79	90%
60	Mistral Large	$0.0067	5.1s	73%	100	100	100	92	86	86	86	86	79	79	90%
49	GPT-4.1	$0.0044	5.8s	77%	100	100	86	86	86	86	86	86	86	86	89%
31	Gemini 2.5 Flash	$0.0013	1.6s	81%	94	92	92	92	92	86	86	86	86	79	89%
33	GPT-4.1 Mini	$0.0010	4.0s	82%	92	92	92	92	92	86	86	86	86	79	89%
48	Gemma 3 27B	$0.0003	8.1s	73%	100	100	92	86	86	86	86	86	86	72	88%
51	Claude 3.5 Haiku	$0.0028	4.4s	73%	94	94	94	94	89	89	89	89	83	64	88%
50	DeepSeek-V2 Chat	$0.0010	7.9s	73%	100	100	94	89	86	86	86	79	79	79	88%
68	DeepSeek V3.2	$0.0008	12.8s	68%	100	100	100	92	92	92	92	79	72	58	88%
87	Nemotron 3 Super	$0.0000	1.0m	82%	94	86	86	86	86	86	86	86	86	86	87%
89	GPT-5.4 Nano (Reasoning)	$0.0018	7.4s	40%	100	100	100	100	100	94	92	92	86	0	87%
63	Claude Sonnet 4.6	$0.013	5.4s	86%	86	86	86	86	86	86	86	86	86	86	86%
39	Claude Haiku 4.5	$0.0042	3.6s	86%	86	86	86	86	86	86	86	86	86	86	86%
58	Mistral Small 3.2 24B	$0.0003	5.2s	68%	100	94	94	92	86	86	86	74	72	66	85%
47	Arcee AI: Trinity Mini	$0.0002	5.2s	74%	94	94	89	83	83	83	83	83	78	78	85%
46	GPT-4o Mini (temp=1)	$0.0003	4.9s	75%	89	89	89	89	83	83	83	83	83	72	84%
90	Llama 3.1 70B	$0.0012	6.9s	41%	94	94	94	94	94	94	94	94	86	0	84%
45	Mistral Large 3	$0.0017	5.3s	79%	92	86	86	86	86	86	79	79	79	79	84%
110	Nemotron 3 Nano	$0.0016	1.8m	69%	94	94	89	86	86	86	86	79	78	58	84%
57	DeepSeek V3 (2024-12-26)	$0.0011	8.7s	73%	94	92	89	86	86	86	79	79	72	72	84%
82	Hermes 3 405B	$0.0025	12.1s	60%	100	94	89	86	86	86	86	86	79	40	83%
36	GPT-4o Mini (temp=0)	$0.0003	3.4s	83%	83	83	83	83	83	83	83	83	83	83	83%
53	Grok 4.20 (Beta)	$0.0033	1.5s	75%	92	86	86	86	86	86	81	79	72	72	83%
73	DeepSeek V3.1	$0.0008	14.6s	69%	94	92	92	86	86	86	79	79	66	64	83%
74	Mistral Large 2	$0.0067	5.3s	70%	100	86	86	86	86	79	79	79	72	72	83%
92	GPT-4o, May 13th (temp=1)	$0.014	2.1s	59%	100	94	94	94	94	89	89	56	56	56	82%
65	Gemini 2.5 Flash Lite	$0.0002	1.2s	66%	94	94	89	86	79	79	79	79	72	66	82%
64	Gemma 3 12B	$0.0001	9.3s	73%	92	86	86	86	86	86	79	79	72	64	82%
55	GPT-5.4 Mini	$0.0015	1.1s	73%	89	86	83	83	83	83	83	83	79	64	82%
70	Qwen 2.5 72B	$0.0005	5.9s	68%	94	89	89	89	89	89	83	64	64	64	81%
52	Ministral 3 8B	$0.0004	3.0s	75%	86	86	86	81	81	81	81	81	76	76	81%
77	Arcee AI: Trinity Large (Preview)	$0.0000	10.4s	65%	89	86	86	83	79	72	72	72	72	66	78%
75	Ministral 8B	$0.0003	3.4s	62%	100	86	81	81	81	76	71	69	69	64	78%
98	WizardLM 2 8x22b	$0.0018	5.9s	38%	100	94	92	86	86	79	79	72	72	0	76%
107	Hermes 3 70B	$0.0008	53.3s	54%	86	81	79	79	79	72	72	72	58	33	71%
112	Claude 3.7 Sonnet	$0.013	6.0s	12%	100	100	100	100	100	100	92	3	3	3	70%
100	Cohere Command R+ (Aug. 2024)	$0.0081	6.4s	47%	89	89	86	79	79	66	64	61	49	33	70%
93	Claude 3 Haiku	$0.0010	3.5s	53%	79	79	79	79	66	64	58	58	58	58	68%
91	Mistral NeMO	$0.0004	2.7s	54%	79	79	72	72	64	64	64	58	58	53	67%
88	Ministral 3 14B	$0.0005	3.3s	58%	72	72	72	72	66	66	64	61	58	53	66%
102	Mistral Small 4	$0.0005	3.4s	39%	86	72	69	66	61	52	49	44	43	33	58%
105	Ministral 3 3B	$0.0003	2.8s	37%	86	56	53	52	52	49	41	41	39	37	51%
106	Ministral 3B	$0.0001	2.4s	36%	72	66	61	60	49	49	40	38	33	33	50%
108	Gemma 3 4B	$0.0001	3.0s	38%	64	58	56	49	49	44	42	39	33	33	47%
113	Rocinante 12B	$0.0005	3.7s	18%	72	72	64	64	51	33	33	3	0	0	39%
114	Llama 3.1 8B	$0.0001	10.6s	27%	44	44	42	40	39	38	33	33	3	3	32%
115	GPT-5.4 Nano (Reasoning, Low)	$0.0009	3.2s	6%	86	60	56	52	28	3	3	0	0	0	29%
116	GPT-5.4 Nano	$0.0005	1.8s	1%	66	53	49	27	3	0	0	0	0	0	20%
117	GPT-4.1 Nano	$0.0002	1.9s	0%	56	33	17	17	0	0	0	0	0	0	12%
118	LFM2 24B	$0.0001	594ms	0%	0	0	0	0	0	0	0	0	0	0	0%
86.71%

Median	Evaluator	Top 3	Flop 3
86.7%	Accuracy (recall)	100Claude Opus 4.6 100ByteDance Seed 2.0 Mini 100MoonshotAI: Kimi K2.5	0Llama 3.1 8B 0LFM2 24B 2GPT-4.1 Nano
93.4%	Precision	100Z.AI GLM 5 Turbo 100Qwen 3.5 397B A17B 100GPT-4o, May 13th (temp=0)	0LFM2 24B 10GPT-4.1 Nano 15Llama 3.1 8B
100.0%	Structural validity	100ByteDance Seed 1.6 100Z.AI GLM 5 Turbo 100Gemini 3 Flash (Preview)	0LFM2 24B 25GPT-4.1 Nano 36GPT-5.4 Nano

Codex Violation Detection

10 codex entries

Performance Score Distribution (Top 20)

Price-Performance Score Distribution (Top 20)

Most Stable Models (Top 20)

Top Overall Models (Top 20)