Value ranking

Best Value LLMs in October 2026

As of October 2026, Qwen3.7 Flash gives the most performance per dollar: 39.9 points at a blended $0.06 per million tokens. Nemotron 3 Nano 30B A3B is next.

Last verified

Noometry Index points per dollar of blended API price, among models scoring in the top half.

Best Value LLMs: top 10
  1. Qwen3.7 Flash 726.0 pts/$
  2. Nemotron 3 Nano 30B A3B 463.5 pts/$
  3. Nemotron 3.5 Lightning 457.5 pts/$
  4. Gemma 4 26B A4B IT 406.6 pts/$
  5. Granite 4.2 8B 377.1 pts/$
  6. Hy3 306.1 pts/$
  7. Gemma 4 31B IT 285.5 pts/$
  8. Step 3.5 Flash 281.9 pts/$
  9. MiMo-V2.6-Flash 277.4 pts/$
  10. GPT-6 Luna 266.5 pts/$

Full ranking

Best Value LLMs
#ModelScoreInput $/MOutput $/MPoints per $
1Qwen3.7 Flash39.9$0.03$0.13726.0
2Nemotron 3 Nano 30B A3B40.6$0.05$0.20463.5
3Nemotron 3.5 Lightning40.0$0.05$0.20457.5
4Gemma 4 26B A4B IT43.5$0.0675$0.23406.6
5Granite 4.2 8B40.5$0.06$0.25377.1
6Hy344.2$0.0825$0.33306.1
7Gemma 4 31B IT43.5$0.09$0.34285.5
8Step 3.5 Flash42.3$0.10$0.30281.9
9MiMo-V2.6-Flash48.5$0.14$0.28277.4
10GPT-6 Luna53.3$0.10$0.50266.5
11MiMo-V2-Omni43.6$0.14$0.28249.1
12MiMo-V2.543.4$0.14$0.28248.2
13Claude Haiku 5.549.5$0.10$0.50247.3
14Qwen3.5-Flash42.5$0.10$0.40243.1
15MiMo-V2-Flash41.3$0.14$0.28236.1
16Nemotron 3 Super40.1$0.08$0.45232.6
17GLM-5.3-Flash51.8$0.15$0.50218.3
18DeepSeek V4 Flash53.6$0.15$0.60204.2
19DeepSeek V4.1 Flash52.8$0.15$0.60201.0
20DeepSeek-V3.2-Exp44.3$0.26$0.38152.6
21Grok 4.1 Fast41.4$0.20$0.50150.4
22GPT-5.6 Luna54.6$0.20$1.20121.3
23Mercury 239.1$0.25$0.75104.3
24Mistral Large 339.1$0.25$0.75104.3
25DeepSeek-V3.142.8$0.25$0.95100.8
26Nvidia Llama 3.3 Nemotron Super 49b v1.540.3$0.40$0.40100.7
27DeepSeek-V339.5$0.24$0.9097.5
28DeepSeek-V3.1-Terminus43.1$0.27$195.4
29MiMo-V2.6-Pro50.3$0.43$0.8792.5
30GLM-4.6V41.3$0.30$0.9091.8
31GLM-4.5-Air38.9$0.20$1.1091.6
32GPT-5.4 nano41.9$0.20$1.2590.6
33MiniMax-M343.8$0.30$1.2083.4
34MiMo-V2.5-Pro45.2$0.43$0.8783.2
35Solar Pro442.1$0.30$1.2080.2
36MiMo-V2-Pro43.0$0.43$0.8779.0
37MiniMax-M2.138.9$0.30$1.2074.2
38Inkling-Small46.5$0.45$1.2072.9
39Gemini 3.1 Flash Lite40.8$0.25$1.5072.5
40Qwen3.7 Plus45.3$0.40$1.6064.8
41Qwen3.5 35B-A3B42.0$0.25$261.0
42GPT-5 Mini41.8$0.25$260.9
43DeepSeek V4 Pro54.3$0.66$1.9854.9
44Kimi K2.548.1$0.45$2.2553.4
45Qwen3.5 27B41.9$0.30$2.4050.7
46Qwen3-Next 80B-A3B Instruct43.0$0.50$249.2
47Gemini 3.5 Flash Lite41.5$0.30$2.5048.8
48Qwen3.5 Plus42.9$0.40$2.4047.6
49Nova 2 Lite39.7$0.30$2.5046.7
50Gemini 3 Flash Preview52.3$0.50$346.5
51DeepSeek-R142.3$0.50$2.1546.4
52DeepSeek-V3.2-Speciale39.7$0.58$1.6846.4
53Gemini 2.5 Flash39.3$0.30$2.5046.2
54Nemotron 3 Ultra42.5$0.50$2.2045.9
55GLM-4.5V39.8$0.60$1.8044.2
56Qwen3.6 Plus47.5$0.50$342.2
57GLM-4.542.0$0.60$2.2042.0
58GLM-4.742.0$0.60$2.2042.0
59MiniMax M140.3$0.55$2.2041.9
60Mistral Large 443.1$0.68$2.0941.7
61GLM-4.641.4$0.60$2.2041.4
62Qwen3.8 27B46.0$0.99$1.4941.2
63Gemini 3.8 Flash61.8$0.75$3.7541.2
64Kimi K2 (Jul 2025)41.2$0.57$2.3041.1
65Hy4 preview45.3$0.75$2.2540.2
66Gemini 3.7 Flash59.8$0.75$3.7539.9
67Seed 2.0 Pro43.2$0.50$338.4
68Qwen3.5 122B-A10B42.1$0.40$3.2038.3
69Gemini 3.6 Flash54.1$0.75$3.7536.1
70Qwen3 235B-A22B43.5$0.70$2.8035.5
71Qwen3-VL 235B-A22B43.2$0.70$2.8035.3
72Qwen3.5 397B-A17B46.0$0.60$3.6034.1
73Step 5 Preview47.9$1$2.7033.6
74Qwen3 32B39.2$0.70$2.8032.0
75Qwen3.6 27B42.2$0.60$3.6031.3
76Grok 4.20 (Non-Reasoning)48.6$1.25$2.5031.1
77GLM-546.1$1$3.2029.7
78Grok 4.20 Multi-Agent46.2$1.25$2.5029.6
79Grok 4.343.8$1.25$2.5028.0
80Kimi K2.647.7$0.95$427.8
81Muse Spark 1.354.8$1.25$4.2527.4
82GPT-5.4 mini45.0$0.75$4.5026.6
83GLM-5.354.8$1.40$4.4025.5
84Kimi K2.7 Code43.3$0.95$425.3
85Muse Spark 1.250.3$1.25$4.2525.2
86Muse Spark 1.149.9$1.25$4.2525.0
87GLM-5.251.1$1.40$4.4023.8
88GLM-5V-Turbo43.8$1.20$423.0
89GLM-5.147.8$1.40$4.4022.2
90o4-mini41.6$1.10$4.4021.6
91Claude Haiku 4.539.5$1$519.8
92Grok 4.656.9$2$619.0
93Qwen3.8 Max56.8$2$618.9
94Grok 4.555.0$2$618.3
95Qwen3 Max43.7$1.20$618.2
96Grok 4.753.1$2$617.7
97Qwen3.6 Max Preview51.5$1.30$7.8017.6
98Inkling44.1$1.87$4.6817.1
99GPT-6.1 Sol65.6$2$1016.4
100Gemini 3.5 Flash54.2$1.50$916.1
101Claude Sonnet 5.561.9$2$1015.5
102GPT-6 Sol61.8$2$1015.4
103GPT-550.9$1.25$1014.8
104GPT-5.149.0$1.25$1014.3
105Qwen3.7 Max51.5$2.50$7.5013.7
106Claude Sonnet 554.6$2$1013.7
107o347.5$2$813.6
108Mistral Medium 3.540.2$1.50$7.5013.4
109GPT-5.6 Terra59.2$2$1213.1
110Gemini 2.5 Pro45.0$1.25$1013.1
111Gemini 3.1 Pro Preview56.7$2$1212.6
112GPT-5.254.1$1.75$1411.2
113GPT-5.459.4$2.50$1510.6
114Kimi K359.5$3$159.9
115GPT-5.3 Codex45.8$1.75$149.5
116GPT-5.3 Chat42.8$1.75$148.9
117GPT-5.2 Codex42.6$1.75$148.8
118Claude Opus 5.568.6$4$208.6
119Claude Sonnet 4.650.3$3$158.4
120GPT-5.6 Sol65.0$4$208.1
121Claude Sonnet 4.544.1$3$157.3
122Claude Sonnet 440.8$3$156.8
123Claude Opus 567.8$5$256.8
124Claude Opus 4.860.7$5$256.1
125Claude Opus 4.758.3$5$255.8
126Claude Opus 4.658.2$5$255.8
127GPT-5.563.4$5$305.6
128Claude Opus 4.550.5$5$255.0
129GPT-6 Astra70.8$10$503.5
130Claude Fable 5.169.0$10$503.5
131Claude Fable 566.8$10$503.3
132o140.9$15$601.6
133Claude Opus 443.1$15$751.4
134Claude Opus 4.141.0$15$751.4
135o3-pro42.9$20$801.2
136GPT-5 Pro46.4$15$1201.1
137GPT-5.5 Pro64.3$30$1801.0
138GPT-5.2 Pro52.3$21$1680.9
139GPT-5.4 Pro58.9$30$1800.9

Sponsored placements are available on pages like this one. Advertise on Noometry

Compare the leaders

Related rankings

Frequently asked questions

Which AI model gives the most performance per dollar?

As of October 2026, Qwen3.7 Flash gives the most performance per dollar: 39.9 points at a blended $0.06 per million tokens. Nemotron 3 Nano 30B A3B is next.

What are the top 5 in this ranking?

Qwen3.7 Flash, Nemotron 3 Nano 30B A3B, Nemotron 3.5 Lightning, Gemma 4 26B A4B IT, Granite 4.2 8B, in that order, as of October 2026.

How is this list ranked?

By score per dollar: the model’s score divided by its blended API price (three parts input to one part output), among models in the top half by score.