LMC ValueScore is a tilted Cobb-Douglas index that weights benchmark quality 60 percent and blended price efficiency 40 percent on a 0-100 scale. It replaces naive quality-divided-by-price rankings that implode near free tiers. 316 scored models, updated hourly. Read the full methodology.
Blended price uses 85 percent input and 15 percent output weighting, reflecting real chat and RAG workload patterns. Reasoning models have their output contribution multiplied by a family-specific expansion factor.
| Rank | Model | LMC ValueScore | Quality | Blended $/M |
|---|---|---|---|---|
| 1 | DeepSeek V4 FlashDeepSeek | 84 | 77 | $0.104 |
| 2 | Gemma 4 31BGoogle | 82 | 80 | $0.155 |
| 3 | Gemini 2.5 Flash Lite Preview 09-2025Google | 82 | 79 | $0.145 |
| 4 | Gemini 2.5 Flash LiteGoogle | 82 | 79 | $0.145 |
| 5 | Gemma 4 31B (free)Google | 81 | 80 | Free |
| 6 | Gemma 4 26B A4B Google | 81 | 73 | $0.101 |
| 7 | MiMo-V2.5Xiaomi | 79 | 72 | $0.131 |
| 8 | DeepSeek V3.2DeepSeek | 78 | 81 | $0.246 |
| 9 | Qwen3.5-FlashAlibaba | 78 | 68 | $0.094 |
| 10 | Hy3 previewTencent | 78 | 68 | $0.085 |
| 11 | Gemma 4 26B A4B (free)Google | 76 | 73 | Free |
| 12 | MiniMax M2.5MiniMax | 75 | 78 | $0.262 |
| 13 | Step 3.5 FlashStepFun | 75 | 67 | $0.122 |
| 14 | Qwen3.5-9BAlibaba | 75 | 66 | $0.107 |
| 15 | Qwen3 235B A22B Thinking 2507Alibaba | 75 | 65 | $0.100 |
| 16 | Qwen3.5-35B-A3BAlibaba | 74 | 76 | $0.269 |
| 17 | Llama 3.3 70B InstructMeta | 74 | 66 | $0.133 |
| 18 | Qwen3 235B A22B Instruct 2507Alibaba | 74 | 64 | $0.092 |
| 19 | GPT-5.1-Codex-MiniOpenAI | 73 | 87 | $0.513 |
| 20 | DeepSeek V4 ProDeepSeek | 73 | 86 | $0.500 |
| 21 | GPT-5.4 NanoOpenAI | 73 | 79 | $0.358 |
| 22 | GLM 4.5 AirZhipu AI | 72 | 70 | $0.238 |
| 23 | Qwen3 Next 80B A3B Instruct (free)Alibaba | 72 | 67 | Free |
| 24 | GLM 4.7 FlashZhipu AI | 72 | 63 | $0.111 |
| 25 | Gemini 3.1 Flash Lite PreviewGoogle | 71 | 79 | $0.438 |
| 26 | DeepSeek V3 0324DeepSeek | 71 | 71 | $0.286 |
| 27 | GPT-4o-miniOpenAI | 71 | 69 | $0.218 |
| 28 | Qwen3 30B A3B Thinking 2507Alibaba | 71 | 64 | $0.128 |
| 29 | Qwen3.5-27BAlibaba | 70 | 77 | $0.400 |
| 30 | Llama 4 MaverickMeta | 70 | 67 | $0.218 |
| 31 | Llama 3.3 70B Instruct (free)Meta | 70 | 66 | Free |
| 32 | DeepSeek V3.2 ExpDeepSeek | 69 | 70 | $0.291 |
| 33 | DeepSeek V3.1DeepSeek | 69 | 69 | $0.297 |
| 34 | DeepSeek V3DeepSeek | 69 | 69 | $0.290 |
| 35 | Qwen3 Next 80B A3B InstructAlibaba | 69 | 67 | $0.242 |
| 36 | MiniMax M3MiniMax | 68 | 74 | $0.435 |
| 37 | MiniMax M2MiniMax | 68 | 72 | $0.367 |
| 38 | Qwen3 VL 235B A22B InstructAlibaba | 68 | 69 | $0.302 |
| 39 | Qwen3 30B A3BAlibaba | 68 | 64 | $0.177 |
| 40 | Qwen3 8BAlibaba | 68 | 61 | $0.102 |
| 41 | Phi 4Microsoft | 68 | 60 | $0.081 |
| 42 | Qwen3.5-122B-A10BAlibaba | 67 | 77 | $0.533 |
| 43 | MiMo-V2.5-ProXiaomi | 67 | 76 | $0.500 |
| 44 | MiniMax M2.7MiniMax | 67 | 69 | $0.348 |
| 45 | Qwen3 Next 80B A3B ThinkingAlibaba | 67 | 64 | $0.200 |
| 46 | Gemini 3 Flash PreviewGoogle | 66 | 88 | $0.875 |
| 47 | Gemini 2.5 FlashGoogle | 66 | 79 | $0.630 |
| 48 | MiniMax M2.1MiniMax | 66 | 70 | $0.389 |
| 49 | DeepSeek V3.1 TerminusDeepSeek | 66 | 69 | $0.372 |
| 50 | Qwen3.5 397B A17BAlibaba | 65 | 79 | $0.695 |
| 51 | Gemma 2 27BGoogle | 65 | 77 | $0.650 |
| 52 | Qwen3.6 PlusAlibaba | 65 | 74 | $0.569 |
| 53 | MiniMax M2-herMiniMax | 64 | 69 | $0.435 |
| 54 | GLM 4.7Zhipu AI | 63 | 72 | $0.603 |
| 55 | GLM 4.6VZhipu AI | 63 | 65 | $0.390 |
| 56 | GLM 5Zhipu AI | 62 | 78 | $0.798 |
| 57 | Llama 3.1 70B InstructMeta | 62 | 65 | $0.400 |
| 58 | Trinity Large Thinkingarcee-ai | 62 | 63 | $0.332 |
| 59 | GPT-4o-mini Search PreviewOpenAI | 62 | 60 | $0.218 |
| 60 | MiniMax M1MiniMax | 61 | 70 | $0.670 |
| 61 | GLM 4.6Zhipu AI | 61 | 70 | $0.627 |
| 62 | gpt-oss-20bOpenAI | 61 | 57 | $0.046 |
| 63 | GLM 4.5Zhipu AI | 60 | 75 | $0.840 |
| 64 | Qwen3 VL 235B A22B ThinkingAlibaba | 60 | 67 | $0.611 |
| 65 | Mercury 2Inception | 59 | 61 | $0.325 |
| 66 | Mistral Large 3 2512Mistral AI | 58 | 67 | $0.650 |
| 67 | GPT-5 MiniOpenAI | 58 | 64 | $0.513 |
| 68 | Grok 4.20xAI | 57 | 88 | $1.44 |
| 69 | Grok 4.20 Multi-AgentxAI | 57 | 87 | $1.44 |
| 70 | Kimi K2.6Moonshot AI | 57 | 75 | $1.07 |
| 71 | gpt-oss-20b (free)OpenAI | 57 | 57 | Free |
| 72 | GLM 5.2Zhipu AI | 56 | 78 | $1.26 |
| 73 | Llama 3.3 Nemotron Super 49B V1.5NVIDIA | 56 | 60 | $0.400 |
| 74 | GPT-5.4 MiniOpenAI | 55 | 79 | $1.31 |
| 75 | Composer 2Cursor | 55 | 65 | $0.800 |
| 76 | Grok 4.3xAI | 54 | 81 | $1.44 |
| 77 | GLM 5.1Zhipu AI | 54 | 76 | $1.29 |
| 78 | Granite 4.1 8BIBM | 54 | 55 | $0.057 |
| 79 | GLM 4.5VZhipu AI | 52 | 62 | $0.780 |
| 80 | Nova 2 LiteAmazon | 52 | 60 | $0.630 |
| 81 | GPT-4o-mini (2024-07-18)OpenAI | 52 | 56 | $0.218 |
| 82 | Llama 4 ScoutMeta | 52 | 55 | $0.130 |
| 83 | Qwen3 Max ThinkingAlibaba | 51 | 68 | $1.25 |
| 84 | Qwen3 MaxAlibaba | 50 | 67 | $1.25 |
| 85 | Kimi K2.5Moonshot AI | 50 | 59 | $0.622 |
| 86 | GLM 5 TurboZhipu AI | 48 | 71 | $1.62 |
| 87 | Olmo 3 32B ThinkAllen AI | 48 | 55 | $0.203 |
| 88 | Qwen3.6 Max PreviewAlibaba | 47 | 75 | $1.82 |
| 89 | Claude Haiku 4.5Anthropic | 47 | 69 | $1.60 |
| 90 | GPT-5 CodexOpenAI | 44 | 88 | $2.56 |
| 91 | GPT-5OpenAI | 44 | 88 | $2.56 |
| 92 | GPT-5.1-Codex-MaxOpenAI | 44 | 87 | $2.56 |
| 93 | GPT-5.1OpenAI | 44 | 87 | $2.56 |
| 94 | GPT-5.1-CodexOpenAI | 44 | 87 | $2.56 |
| 95 | GPT-5.1 ChatOpenAI | 44 | 86 | $2.56 |
| 96 | Gemini 2.5 ProGoogle | 43 | 83 | $2.56 |
| 97 | Gemini 2.5 Pro Preview 06-05Google | 43 | 83 | $2.56 |
| 98 | Gemini 2.5 Pro Preview 05-06Google | 43 | 83 | $2.56 |
| 99 | Phi 4 Mini InstructMicrosoft | 42 | 52 | $0.121 |
| 100 | Gemini 3.5 FlashGoogle | 41 | 79 | $2.63 |
| 101 | Mistral Medium 3.5Mistral AI | 40 | 71 | $2.40 |
| 102 | GPT-5 ChatOpenAI | 38 | 70 | $2.56 |
| 103 | Composer 2 FastCursor | 38 | 65 | $2.40 |
| 104 | R1 0528DeepSeek | 37 | 79 | $3.00 |
| 105 | Qwen3 235B A22BAlibaba | 37 | 54 | $0.660 |
| 106 | Mistral LargeMistral AI | 36 | 66 | $2.60 |
| 107 | GPT-4.1 MiniOpenAI | 36 | 53 | $0.580 |
| 108 | Gemini 3.1 Pro Preview Custom ToolsGoogle | 34 | 92 | $3.50 |
| 109 | Gemini 3.1 Pro PreviewGoogle | 34 | 92 | $3.50 |
| 110 | GPT-4.1OpenAI | 34 | 67 | $2.90 |
| 111 | Mixtral 8x22B InstructMistral AI | 34 | 63 | $2.60 |
| 112 | GPT-5.2-CodexOpenAI | 33 | 90 | $3.59 |
| 113 | GPT-5.2OpenAI | 33 | 90 | $3.59 |
| 114 | Kimi K2.7 CodeMoonshot AI | 33 | 54 | $1.15 |
| 115 | Kimi K2 ThinkingMoonshot AI | 33 | 53 | $0.885 |
| 116 | GPT-5.3-CodexOpenAI | 32 | 88 | $3.59 |
| 117 | GPT-5.3 ChatOpenAI | 32 | 87 | $3.59 |
| 118 | Claude 3 HaikuAnthropic | 31 | 51 | $0.400 |
| 119 | GPT-5.2 ChatOpenAI | 30 | 77 | $3.59 |
| 120 | Kimi K2 0905Moonshot AI | 30 | 52 | $0.885 |
| 121 | R1DeepSeek | 29 | 74 | $3.60 |
| 122 | GPT-4o (2024-08-06)OpenAI | 28 | 71 | $3.63 |
| 123 | GPT-4oOpenAI | 28 | 71 | $3.63 |
| 124 | GPT-4o Search PreviewOpenAI | 28 | 70 | $3.63 |
| 125 | Mistral Large 2407Mistral AI | 27 | 56 | $2.60 |
| 126 | Kimi K2 0711Moonshot AI | 27 | 51 | $0.830 |
| 127 | Command R (08-2024)Cohere | 23 | 48 | $0.218 |
| 128 | o4 MiniOpenAI | 22 | 81 | $4.24 |
| 129 | GPT-5.4OpenAI | 21 | 92 | $4.38 |
| 130 | o4 Mini HighOpenAI | 20 | 72 | $4.24 |
| 131 | GPT-5 NanoOpenAI | 18 | 46 | $0.102 |
| 132 | Claude Fable 5Anthropic | 16 | 97 | $16.00 |
| 133 | Claude Opus 4.7 (Fast)Anthropic | 15 | 95 | $48.00 |
| 134 | Claude Opus 4.7Anthropic | 15 | 95 | $8.00 |
| 135 | Claude Opus 4.8 (Fast)Anthropic | 15 | 94 | $16.00 |
| 136 | Claude Opus 4.8Anthropic | 15 | 94 | $8.00 |
| 137 | GPT-5.5OpenAI | 15 | 92 | $8.75 |
| 138 | GPT-5.4 ProOpenAI | 15 | 92 | $52.50 |
| 139 | GPT-5.5 ProOpenAI | 15 | 90 | $52.50 |
| 140 | GPT-5.2 ProOpenAI | 15 | 90 | $43.05 |
| 141 | Claude Opus 4.6 (Fast)Anthropic | 15 | 90 | $48.00 |
| 142 | Claude Opus 4.6Anthropic | 15 | 90 | $8.00 |
| 143 | GPT-5 ProOpenAI | 15 | 88 | $30.75 |
| 144 | o3 Deep ResearchOpenAI | 15 | 86 | $68.50 |
| 145 | o3 ProOpenAI | 15 | 86 | $137.00 |
| 146 | o3OpenAI | 15 | 86 | $13.70 |
| 147 | GPT-4o (2024-11-20)OpenAI | 15 | 53 | $3.63 |
| 148 | Claude Sonnet 4.6Anthropic | 14 | 85 | $4.80 |
| 149 | Claude Opus 4.5Anthropic | 14 | 85 | $8.00 |
| 150 | Claude Sonnet 4.5Anthropic | 14 | 82 | $4.80 |
| 151 | Claude Opus 4Anthropic | 14 | 82 | $24.00 |
| 152 | o4 Mini Deep ResearchOpenAI | 14 | 81 | $7.70 |
| 153 | o3 MiniOpenAI | 13 | 75 | $7.54 |
| 154 | Claude Opus 4.1Anthropic | 13 | 75 | $24.00 |
| 155 | Claude Sonnet 4Anthropic | 13 | 74 | $4.80 |
| 156 | o1OpenAI | 13 | 74 | $84.75 |
| 157 | o1-proOpenAI | 13 | 73 | $1477.50 |
| 158 | GPT-4o (2024-05-13)OpenAI | 13 | 71 | $6.50 |
| 159 | GPT-4 TurboOpenAI | 12 | 66 | $13.00 |
| 160 | GPT-4OpenAI | 12 | 65 | $34.50 |
| 161 | o3 Mini HighOpenAI | 12 | 64 | $7.54 |
| 162 | Command ACohere | 12 | 50 | $3.63 |
| 163 | GPT-4 Turbo PreviewOpenAI | 11 | 59 | $13.00 |
| 164 | Devstral 2 2512Mistral AI | 11 | 45 | $0.640 |
| 165 | Llama 3.1 8B InstructMeta | 11 | 44 | $0.022 |
| 166 | Command R (08-2024)Cohere | 9 | 48 | $3.63 |
| 167 | GPT-4.1 NanoOpenAI | 6 | 42 | $0.145 |
| 168 | gpt-oss-120bOpenAI | 4 | 40 | $0.060 |
| 169 | North Mini Code (free)Cohere | 4 | 40 | Free |
| 170 | Nemotron 3.5 Content Safety (free)NVIDIA | 4 | 40 | Free |
| 171 | Nemotron 3 Ultra (free)NVIDIA | 4 | 40 | Free |
| 172 | Qwen3.7 PlusAlibaba | 4 | 40 | $0.464 |
| 173 | Step 3.7 FlashStepFun | 4 | 40 | $0.342 |
| 174 | Perceptron Mk1perceptron | 4 | 40 | $0.353 |
| 175 | Ring-2.6-1Tinclusionai | 4 | 40 | $0.158 |
| 176 | Nemotron 3 Nano Omni (free)NVIDIA | 4 | 40 | Free |
| 177 | Laguna XS.2 (free)poolside | 4 | 40 | Free |
| 178 | Laguna XS.2poolside | 4 | 40 | $0.115 |
| 179 | Laguna M.1 (free)poolside | 4 | 40 | Free |
| 180 | Laguna M.1poolside | 4 | 40 | $0.230 |
| 181 | Qwen3.6 FlashAlibaba | 4 | 40 | $0.328 |
| 182 | Qwen3.6 35B A3BAlibaba | 4 | 40 | $0.269 |
| 183 | Ling-2.6-1Tinclusionai | 4 | 40 | $0.158 |
| 184 | Ling-2.6-flashinclusionai | 4 | 40 | $0.013 |
| 185 | Lyria 3 Pro PreviewGoogle | 4 | 40 | Free |
| 186 | Lyria 3 Clip PreviewGoogle | 4 | 40 | Free |
| 187 | KAT-Coder-Pro V2Kuaishou | 4 | 40 | $0.435 |
| 188 | Reka Edgerekaai | 4 | 40 | $0.100 |
| 189 | Mistral Small 4Mistral AI | 4 | 40 | $0.218 |
| 190 | Nemotron 3 Super (free)NVIDIA | 4 | 40 | Free |
| 191 | Nemotron 3 SuperNVIDIA | 4 | 40 | $0.144 |
| 192 | Seed-2.0-MiniByteDance | 4 | 40 | $0.145 |
| 193 | LFM2-24B-A2BLiquid AI | 4 | 40 | $0.044 |
| 194 | Qwen3.5 Plus 2026-02-15Alibaba | 4 | 40 | $0.455 |
| 195 | Qwen3 Coder NextAlibaba | 4 | 40 | $0.213 |
| 196 | Solar Pro 3Upstage | 4 | 40 | $0.218 |
| 197 | LFM2.5-1.2B-Thinking (free)Liquid AI | 4 | 40 | Free |
| 198 | LFM2.5-1.2B-Instruct (free)Liquid AI | 4 | 40 | Free |
| 199 | Seed 1.6 FlashByteDance | 4 | 40 | $0.109 |
| 200 | Nemotron 3 Nano 30B A3B (free)NVIDIA | 4 | 40 | Free |
| 201 | Nemotron 3 Nano 30B A3BNVIDIA | 4 | 40 | $0.073 |
| 202 | Ministral 3 14B 2512Mistral AI | 4 | 40 | $0.200 |
| 203 | Ministral 3 8B 2512Mistral AI | 4 | 40 | $0.150 |
| 204 | Ministral 3 3B 2512Mistral AI | 4 | 40 | $0.100 |
| 205 | Trinity Miniarcee-ai | 4 | 40 | $0.061 |
| 206 | Voxtral Small 24B 2507Mistral AI | 4 | 40 | $0.130 |
| 207 | gpt-oss-safeguard-20bOpenAI | 4 | 40 | $0.109 |
| 208 | Nemotron Nano 12B 2 VL (free)NVIDIA | 4 | 40 | Free |
| 209 | Qwen3 VL 32B InstructAlibaba | 4 | 40 | $0.151 |
| 210 | Granite 4.0 MicroIBM | 4 | 40 | $0.031 |
| 211 | Qwen3 VL 8B ThinkingAlibaba | 4 | 40 | $0.304 |
| 212 | Qwen3 VL 8B InstructAlibaba | 4 | 40 | $0.143 |
| 213 | Qwen3 VL 30B A3B ThinkingAlibaba | 4 | 40 | $0.345 |
| 214 | Qwen3 VL 30B A3B InstructAlibaba | 4 | 40 | $0.189 |
| 215 | Qwen3 Coder FlashAlibaba | 4 | 40 | $0.312 |
| 216 | Qwen Plus 0728 (thinking)Alibaba | 4 | 40 | $0.338 |
| 217 | Qwen Plus 0728Alibaba | 4 | 40 | $0.338 |
| 218 | Nemotron Nano 9B V2 (free)NVIDIA | 4 | 40 | Free |
| 219 | Codestral 2508Mistral AI | 4 | 40 | $0.390 |
| 220 | Qwen3 Coder 30B A3B InstructAlibaba | 4 | 40 | $0.100 |
| 221 | Qwen3 30B A3B Instruct 2507Alibaba | 4 | 40 | $0.070 |
| 222 | Qwen3 Coder 480B A35B (free)Alibaba | 4 | 40 | Free |
| 223 | Qwen3 Coder 480B A35BAlibaba | 4 | 40 | $0.457 |
| 224 | UI-TARS 7B ByteDance | 4 | 40 | $0.115 |
| 225 | Hunyuan A13B InstructTencent | 4 | 40 | $0.205 |
| 226 | Mistral Small 3.2 24BMistral AI | 4 | 40 | $0.094 |
| 227 | Gemma 3n 4BGoogle | 4 | 40 | $0.069 |
| 228 | Llama Guard 4 12BMeta | 4 | 40 | $0.180 |
| 229 | Qwen3 14BAlibaba | 4 | 40 | $0.121 |
| 230 | Qwen3 32BAlibaba | 4 | 40 | $0.110 |
| 231 | Mistral Small 3.1 24BMistral AI | 4 | 40 | $0.382 |
| 232 | Gemma 3 4BGoogle | 4 | 40 | $0.057 |
| 233 | Gemma 3 12BGoogle | 4 | 40 | $0.065 |
| 234 | Reka Flash 3rekaai | 4 | 40 | $0.115 |
| 235 | Gemma 3 27BGoogle | 4 | 40 | $0.092 |
| 236 | SabaMistral AI | 4 | 40 | $0.260 |
| 237 | Qwen-PlusAlibaba | 4 | 40 | $0.338 |
| 238 | Mistral Small 3Mistral AI | 4 | 40 | $0.054 |
| 239 | MiniMax-01MiniMax | 4 | 40 | $0.335 |
| 240 | Nova Lite 1.0Amazon | 4 | 40 | $0.087 |
| 241 | Nova Micro 1.0Amazon | 4 | 40 | $0.051 |
| 242 | Llama 3.2 11B Vision InstructMeta | 4 | 40 | $0.345 |
| 243 | Qwen2.5 72B InstructAlibaba | 4 | 40 | $0.366 |
| 244 | SWE-1.5Windsurf | 4 | 40 | Free |
| 245 | ALLaM 2 7B InstructHUMAIN | 4 | 40 | Free |
| 246 | ALLaM 34BHUMAIN | 4 | 40 | Free |
| 247 | Falcon-H1-Arabic 34B InstructTII | 4 | 40 | Free |
| 248 | Falcon-H1-Arabic 7B InstructTII | 4 | 40 | Free |
| 249 | Falcon-H1-Arabic 3B InstructTII | 4 | 40 | Free |
| 250 | Falcon Arabic 7B InstructTII | 4 | 40 | Free |
| 251 | Falcon3 10B InstructTII | 4 | 40 | Free |
| 252 | Falcon3 7B InstructTII | 4 | 40 | Free |
| 253 | Mistral NemoMistral AI | 4 | 40 | $0.022 |
| 254 | R1 Distill Llama 70BDeepSeek | 3 | 41 | $1.64 |
| 255 | Nemotron 3 UltraNVIDIA | 3 | 40 | $0.755 |
| 256 | Qwen3.7 MaxAlibaba | 3 | 40 | $1.63 |
| 257 | Grok Build 0.1xAI | 3 | 40 | $1.15 |
| 258 | Anthropic Claude Haiku Latest~anthropic | 3 | 40 | $1.60 |
| 259 | OpenAI GPT Mini Latest~openai | 3 | 40 | $1.31 |
| 260 | MoonshotAI Kimi Latest~moonshotai | 3 | 40 | $1.07 |
| 261 | Qwen3.5 Plus 2026-04-20Alibaba | 3 | 40 | $0.525 |
| 262 | Qwen3.6 27BAlibaba | 3 | 40 | $0.721 |
| 263 | GLM 5V TurboZhipu AI | 3 | 40 | $1.62 |
| 264 | Seed-2.0-LiteByteDance | 3 | 40 | $0.513 |
| 265 | Aion-2.0aion-labs | 3 | 40 | $0.920 |
| 266 | Palmyra X5Writer | 3 | 40 | $1.41 |
| 267 | GPT Audio MiniOpenAI | 3 | 40 | $0.870 |
| 268 | Seed 1.6ByteDance | 3 | 40 | $0.513 |
| 269 | Cogito v2.1 671Bdeepcogito | 3 | 40 | $1.25 |
| 270 | Qwen3 Coder PlusAlibaba | 3 | 40 | $1.04 |
| 271 | Mistral Medium 3.1Mistral AI | 3 | 40 | $0.640 |
| 272 | ERNIE 4.5 VL 424B A47B Baidu | 3 | 40 | $0.545 |
| 273 | Virtuoso Largearcee-ai | 3 | 40 | $0.817 |
| 274 | Aion-1.0-Miniaion-labs | 3 | 40 | $0.805 |
| 275 | Qwen2.5 VL 72B InstructAlibaba | 3 | 40 | $0.830 |
| 276 | SonarPerplexity | 3 | 40 | $1.00 |
| 277 | Nova Pro 1.0Amazon | 3 | 40 | $1.16 |
| 278 | Qwen2.5 Coder 32B InstructAlibaba | 3 | 40 | $0.711 |
| 279 | GPT-3.5 TurboOpenAI | 3 | 40 | $0.650 |
| 280 | Coder Largearcee-ai | 3 | 39 | $0.545 |
| 281 | gpt-oss-120b (free)OpenAI | 3 | 39 | Free |
| 282 | ALLaM 7B Instruct (preview)HUMAIN | 3 | 38 | Free |
| 283 | Qwen2.5 7B InstructAlibaba | 3 | 38 | $0.049 |
| 284 | Google Gemini Pro Latest~google | 2 | 40 | $3.50 |
| 285 | Google Gemini Flash Latest~google | 2 | 40 | $2.63 |
| 286 | GPT AudioOpenAI | 2 | 40 | $3.63 |
| 287 | Jamba Large 1.7AI21 Labs | 2 | 40 | $2.90 |
| 288 | Sonar Reasoning ProPerplexity | 2 | 40 | $2.90 |
| 289 | Sonar Deep ResearchPerplexity | 2 | 40 | $2.90 |
| 290 | GPT-3.5 Turbo 16kOpenAI | 2 | 40 | $3.15 |
| 291 | GPT-3.5 Turbo (older v0613)OpenAI | 2 | 39 | $1.15 |
| 292 | Fugu Ultrasakana | 1 | 40 | $8.75 |
| 293 | Claude Fable Latest~anthropic | 1 | 40 | $16.00 |
| 294 | GPT Chat LatestOpenAI | 1 | 40 | $8.75 |
| 295 | Anthropic Claude Sonnet Latest~anthropic | 1 | 40 | $4.80 |
| 296 | OpenAI GPT Latest~openai | 1 | 40 | $8.75 |
| 297 | Claude Opus Latest~anthropic | 1 | 40 | $8.00 |
| 298 | Nova Premier 1.0Amazon | 1 | 40 | $4.00 |
| 299 | Sonar Pro SearchPerplexity | 1 | 40 | $4.80 |
| 300 | Sonar ProPerplexity | 1 | 40 | $4.80 |
| 301 | Aion-1.0aion-labs | 1 | 40 | $4.60 |
| 302 | Command R7B (12-2024)Cohere | 1 | 35 | $0.054 |
| 303 | ALLaM 1 13B InstructHUMAIN | 1 | 35 | $1.80 |
| 304 | Llama 3 8B InstructMeta | 1 | 34 | $0.140 |
| 305 | Falcon Mamba 7B InstructTII | 1 | 34 | Free |
| 306 | GPT-3.5 Turbo InstructOpenAI | 1 | 34 | $1.57 |
| 307 | Llama 3.2 3B Instruct (free)Meta | 1 | 33 | Free |
| 308 | Llama 3.2 3B InstructMeta | 1 | 33 | $0.094 |
| 309 | Inflection 3 ProductivityInflection | 0 | 32 | $3.63 |
| 310 | Inflection 3 PiInflection | 0 | 32 | $3.63 |
| 311 | autofixer-01Vercel | 0 | 30 | Free |
| 312 | WizardLM-2 8x22BMicrosoft | 0 | 29 | $0.620 |
| 313 | MellumJetBrains | 0 | 24 | Free |
| 314 | Gemini 3.1 Flash LiteGoogle | 0 | 23 | $0.438 |
| 315 | Llama 3.2 1B InstructMeta | 0 | 18 | $0.053 |
| 316 | Mistral Medium 3Mistral AI | 0 | 14 | $0.640 |
A tilted Cobb-Douglas index (quality weighted 60 percent) gated by a smooth production-readiness sigmoid. Price anchors frozen to Q2 2026: p10 = $0.087/M, p90 = $4.80/M.
Optimize for value when running high-volume workloads like classification, summarization, or data extraction where a small accuracy delta does not change the outcome. Every 10 points of blended price savings compounds quickly at scale.
Optimize for quality when accuracy is critical: complex reasoning, code generation for production, or agentic workflows where errors compound. A flagship that costs 10x more can save 100x in debugging and rework.
A naive score/price ratio explodes as price approaches zero, which is why free models would always dominate a simple division. LMC ValueScore uses a tilted Cobb-Douglas utility index (Q^0.6 times P^0.4) with a smooth sigmoid gate at composite=50. This prevents unproven free models from dominating, and the 60/40 tilt keeps quality slightly more important than price efficiency, which matches what production buyers actually care about.
Early versions of the formula sent premium flagships to LMC ValueScore 0 whenever their price passed the empirical p90 anchor. That produced a PR-disastrous "Claude = 0, Gemini = 82" headline without distinguishing a $4.80/M model from a $140/M one. The final v3 formula applies a soft floor of 1 on the price efficiency component so quality still shows through. Claude Opus 4.6 at composite 90 lands around 15 rather than 0, which is honest: it is clearly not a best-value pick at 8 dollars per million tokens, but it is still one of the highest-quality models available.
LMC ValueScore uses empirical p10 and p90 anchors computed from the priced text-model catalog: p10 = $0.087/M and p90 = $4.80/M as of Q2 2026. We deliberately freeze these quarterly rather than recomputing on every refresh. Morningstar and MSCI rating systems use the same approach: unstable anchors produce unstable rankings that churn for reasons unrelated to the underlying models, which is exactly what SEO and user trust reward us for avoiding.
Yes. Reasoning models bill hidden "thinking" tokens on top of the visible output, so the list output price understates real cost. LMC ValueScore multiplies the output-token contribution by a family-specific expansion factor: o1 at 8x, o1-pro at 15x, o3 at 10x, R1 at 8x, QwQ at 6x, and so on. These factors come from the o1 system card, the DeepSeek R1 paper, and published cookbook traces. That is why R1 lands around 29-37 on LMC ValueScore even though its nominal per-token price looks competitive.
Real chat and RAG workloads are input-heavy: you stuff a lot of context (system prompt, user history, retrieved documents) and get relatively short answers back. Anthropic published caching statistics showing median input-to-output ratios around 5:1 to 10:1 for production traffic, and OpenAI DevDay data agrees. A flat average would bias against reasoning models and against any provider with expensive output tokens. 85/15 is the tightest fit to the median production workload we can defend with public data.
DeepSeek V4 Flash from DeepSeek currently leads the rankings with an LMC ValueScore of 84, combining a composite quality score of 77 with a blended price of $0.104 per million tokens. The full ranked table below updates hourly.