Best AI models by Effective Value (𝕍)

Ranked by Effective Value (𝕍) — the single metric that fuses accuracy, speed, cost, and the exponential error-cascade of deep agentic work. 𝕍 rewards models that complete long multi-step chains without human rescue. For production agents (not one-shot chat), 𝕍 is the number that matters most. Shown indexed to the top model = 100 (raw 𝕍 spans a ~2.7M× range).

Ranking 82 models across the meo-tested roster · as of 2026-10-18.

#	Model	Lab	𝕍 index
1	Anthropic: Claude Opus 5.5	anthropic	100.0
2	OpenAI: GPT-6.1 Sol	openai	35.6
3	Anthropic: Claude Fable 5	anthropic	27.5
4	OpenAI: GPT-5.5	openai	23.2
5	Google: Gemini 4 Argon	google	18.2
6	inclusionAI: Ring-2.6-1T	inclusionai	14.8
7	Anthropic: Claude Opus 4.8	anthropic	12.1
8	Google: Gemini 3.5 Flash	google	11.1
9	OpenAI: GPT-5.5 Pro	openai	4.3
10	MoonshotAI: Kimi K3	moonshotai	3.6
11	Google: Gemini 3.1 Pro Preview	google	3.2
12	OpenAI: GPT-5.4 Nano	openai	2.3
13	SpaceXAI: Grok 4.3	x-ai	1.7
14	OpenAI: GPT-5.1-Codex-Mini	openai	1.3
15	NVIDIA: Nemotron 3 Ultra	nvidia	1.2
16	Z.ai: GLM 5.2	z-ai	0.78
17	Qwen: Qwen3.7 Max	qwen	0.65
18	DeepSeek: DeepSeek V4 Pro 0423	deepseek	0.53
19	Google: Gemini 3.1 Flash Lite Preview	google	0.46
20	DeepSeek: DeepSeek V4 Flash 0423	deepseek	0.43
21	Owl Alpha	openrouter	0.37
22	OpenAI: GPT-4.1 Mini	openai	0.26
23	Xiaomi: MiMo-V2.5	xiaomi	0.26
24	OpenAI: GPT-5 Mini	openai	0.25
25	Qwen: Qwen3 VL 32B Instruct	qwen	0.19
26	Nex AGI: Nex-N2-Pro (free)	nex-agi	0.18
27	Xiaomi: MiMo-V2.5-Pro	xiaomi	0.18
28	SpaceXAI: Grok 4.7	x-ai	0.17
29	Mistral: Mistral Large 3 2512	mistralai	0.17
30	Qwen: Qwen3.6 Flash	qwen	0.17
31	Qwen: Qwen3.5-Flash	qwen	0.15
32	Google: Gemini 3.1 Flash Lite	google	0.13
33	MiniMax: MiniMax M3	minimax	0.13
34	Google: Gemini 2.5 Flash Lite	google	0.13
35	Qwen: Qwen3 VL 30B A3B Instruct	qwen	0.13
36	Z.ai: GLM 4.6V	z-ai	0.12
37	Mistral: Ministral 3 14B 2512	mistralai	0.086
38	ByteDance Seed: Seed 1.6 Flash	bytedance-seed	0.080
39	MoonshotAI: Kimi K2.7 Code	moonshotai	0.073
40	Qwen: Qwen3 VL 8B Thinking	qwen	0.072
41	Mistral: Ministral 3 8B 2512	mistralai	0.065
42	Mistral: Mistral Small 4	mistralai	0.065
43	Qwen: Qwen3 VL 235B A22B Instruct	qwen	0.060
44	Perceptron: Perceptron Mk1	perceptron	0.055
45	Qwen: Qwen3.7 Plus	qwen	0.051
46	MoonshotAI: Kimi K2.6	moonshotai	0.050
47	OpenAI: GPT-4.1 Nano	openai	0.044
48	ByteDance Seed: Seed 1.6	bytedance-seed	0.036
49	Z.ai: GLM 5.1	z-ai	0.033
50	Qwen: Qwen3 VL 8B Instruct	qwen	0.028
51	OpenAI: GPT-5 Nano	openai	0.026
52	OpenAI: GPT-4o-mini	openai	0.021
53	Mistral: Ministral 3 3B 2512	mistralai	0.021
54	Mistral: Mistral Small 3.2 24B	mistralai	0.017
55	NVIDIA: Nemotron 3 Nano Omni (free)	nvidia	0.016
56	Baidu: ERNIE 4.5 VL 424B A47B	baidu	0.015
57	Anthropic: Claude 3 Haiku	anthropic	0.014
58	Qwen: Qwen3.5 Plus 2026-02-15	qwen	0.012
59	Amazon: Nova Lite 1.0	amazon	0.012
60	Google: Gemma 3 12B	google	0.009
61	Google: Gemma 4 31B	google	0.009
62	Meta: Llama 4 Scout	meta-llama	0.009
63	Meta: Llama 4 Maverick	meta-llama	0.008
64	Qwen: Qwen3 VL 30B A3B Thinking	qwen	0.008
65	Tencent: Hy3 preview	tencent	0.007
66	Google: Gemma 3 27B	google	0.006
67	Qwen: Qwen2.5 VL 72B Instruct	qwen	0.005
68	StepFun: Step 3.7 Flash	stepfun	0.005
69	MiniMax: MiniMax-01	minimax	0.005
70	Qwen: Qwen3.6 35B A3B	qwen	0.004
71	Perplexity: Sonar	perplexity	0.004
72	Google: Gemma 4 26B A4B	google	0.003
73	Google: Gemma 3 4B	google	0.003
74	NVIDIA: Nemotron Nano 12B 2 VL (free)	nvidia	0.002
75	Arcee AI: Trinity Large Thinking	arcee-ai	<0.001
76	Qwen: Qwen3.5-27B	qwen	<0.001
77	ByteDance Seed: Seed-2.0-Lite	bytedance-seed	<0.001
78	Qwen: Qwen3.6 Plus	qwen	<0.001
79	Qwen: Qwen3.5-35B-A3B	qwen	<0.001
80	Meta: Llama 3.2 11B Vision Instruct	meta-llama	<0.001
81	Qwen: Qwen3.5-9B	qwen	<0.001
82	ByteDance Seed: Seed-2.0-Mini	bytedance-seed	<0.001

← All rankings Methodology & 𝕍 →