The AI tools index that doesn't waste your time.不浪费你时间的 AI 工具索引。
LynxModel

LLM API pricing — cost per point of intelligenceLLM API 价格——每美元能买多少智能

Every tracked model with its API list price and a value metric: value = Intelligence Score ÷ blended price (blended = average of input and output price per 1M tokens). Higher value = more intelligence per dollar. Sorted by value by default; click any column to re-sort. Prices sync from the OpenRouter API daily (refreshed 2026-10-20).

所有被追踪模型的 API 标价与性价比指标:性价比 = 智能评分 ÷ 混合价(混合价 = 输入输出每百万 token 价格的均值)。性价比越高,每美元买到的智能越多。默认按性价比排序,点表头可换列。价格每日从 OpenRouter API 同步(更新于 2026-10-20)。

Find my model帮我选型 Leaderboard排行榜 Head-to-head模型对战

Best value right now: Qwen3.7 Flash (1025), DeepSeek V4 Flash 0423 (611), Hy3 (407.3) (score per blended $).当前性价比前三:Qwen3.7 Flash (1025), DeepSeek V4 Flash 0423 (611), Hy3 (407.3)(每混合美元换来的分数)。

#Model模型Score总分 $ in/1M输入$/1M $ out/1M输出$/1M Blended混合价 Value性价比 Ctx上下文
1Qwen3.7 Flash(on/off开/关)Qwen82$0.030$0.130$0.08010251M
2DeepSeek V4 Flash 0423(max)DeepSeek83.4$0.091$0.182$0.1376111M
3Hy3Tencent84$0.083$0.330$0.206407.3262K
4Phi 4Microsoft38.5$0.070$0.140$0.105366.716K
5Seed-2.0-MiniByteDance Seed77.7$0.100$0.400$0.250310.8262K
6DeepSeek V4 Flash 0731DeepSeek91.1$0.010$0.640$0.325280.31M
7gpt-oss-120bOpenAI64.9$0.150$0.600$0.375173.1131K
8DeepSeek V4 Pro 0423(max)DeepSeek87.4$0.372$0.743$0.558156.81M
9DeepSeek V3.2(on/off开/关)DeepSeek70.9$0.259$0.800$0.530133.9164K
10Mistral Small 4Mistral AI50.2$0.150$0.600$0.375133.9262K
11GPT-5.6 Luna(max)OpenAI84.7$0.200$1.20$0.7001211.1M
12Step 3.7 FlashStepFun81.4$0.200$1.15$0.675120.6262K
13MiMo-V2.5-ProXiaomi75$0.435$0.870$0.653114.91.1M
14Mercury 2Inception57.2$0.250$0.750$0.500114.4128K
15MiniMax M3MiniMax81.7$0.300$1.20$0.750108.91M
16Qwen3.7 Plus(on/off开/关)Qwen83.4$0.320$1.28$0.800104.31M
17GPT-5.6 Luna Pro(max)OpenAI71.5$0.200$1.20$0.700102.11.1M
18Llama 4 MaverickMeta42.5$0.188$0.652$0.420101.21M
19MiniMax M2.7MiniMax75.7$0.300$1.20$0.750100.9205K
20Llama 4 ScoutMeta34.5$0.180$0.590$0.38589.61.3M
21Gemini 3.1 Flash Lite(high)Google58.6$0.250$1.50$0.875671M
22Seed-2.0-LiteByteDance Seed65.3$0.250$2.00$1.1358262K
23Nemotron 3 UltraNVIDIA76.4$0.600$2.40$1.5050.9262K
24Kimi K2 Thinking(always-on常开)Moonshot AI71$0.600$2.50$1.5545.8262K
25Grok 4.20(high)xAI85.1$1.25$2.50$1.8845.42M
26Gemini 3.5 Flash Lite(high)Google63.4$0.300$2.50$1.4045.31M
27Kimi K2.7 CodeMoonshot AI80.1$0.671$3.35$2.0139.8262K
28Grok 4.3(high)xAI74.3$1.25$2.50$1.8839.61M
29Gemini 3.6 Flash(high)Google80.1$0.750$3.75$2.2535.61M
30GLM 5.1(on/off开/关)Z.ai70.1$0.966$3.04$2.0035205K
31Muse Spark 1.2meta83$1.25$4.25$2.7530.21M
32Qwen3.7 Max(on/off开/关)Qwen85.2$1.48$4.42$2.9528.91M
33Muse Spark 1.1meta71.8$1.25$4.25$2.7526.11M
34GLM 5.2(max)Z.ai88.7$0.100$7.00$3.55251M
35Grok 4.5(high)xAI79.5$2.00$6.00$4.0019.9500K
36Nova Pro 1.0Amazon39.1$0.800$3.20$2.0019.6300K
37Claude Haiku 4.5(high)Anthropic54.2$1.00$5.00$3.0018.1200K
38Gemini 3.5 Flash(high)Google83.6$1.50$9.00$5.2515.91M
39Mistral Medium 3.5Mistral AI65.4$1.50$7.50$4.5014.5262K
40Claude Sonnet 5(high)Anthropic80.9$2.00$10$6.0013.51M
41GPT-5.6 Terra(max)OpenAI89.7$2.00$12$7.0012.81.1M
42Gemini 3.1 Pro Preview(high)Google89.6$2.00$12$7.0012.81M
43Mistral LargeMistral AI49.9$2.00$6.00$4.0012.5128K
44Kimi K3(always-on常开)Moonshot AI87.9$0.300$14$7.1512.31M
45Qwen3.8 Max PrimeQwen86.4$4.00$12$8.0010.81M
46Command ACohere50.7$2.50$10$6.258.1256K
47GPT-5.6 Sol(max)OpenAI94.3$4.00$20$127.91.1M
48GPT-5.6 Sol Pro(max)OpenAI88.8$4.00$20$127.41.1M
49Command R+ (08-2024)Cohere42.8$2.50$10$6.256.8128K
50Claude Opus 4.8(max)Anthropic87$5.00$25$155.81M
51Sonar ProPerplexity52.5$3.00$15$9.005.8200K
52Claude Opus 5(max)Anthropic85$5.00$25$155.71M
53GPT-5.5(max)OpenAI82$5.00$30$184.71.1M
54Nova Premier 1.0Amazon30.6$2.50$13$7.504.11M
55Claude Fable 5(max)Anthropic95.4$10$50$303.21M

Models without a public score or without list prices are omitted from the value table — see the leaderboard for the full catalog.未出分或无价模型不进性价比表——完整目录见排行榜。

Monthly cost calculator月成本计算器

Estimate each model at your own usage, then set the minimum score you can accept.按你的实际用量估算每个模型的月账单,再设定能接受的最低分。

How the score works →评分方法 → · List prices via the official OpenRouter API; volume discounts, cached-token pricing and batch APIs not included.价格为 OpenRouter 官方 API 标价,未含阶梯折扣、缓存命中价与批量 API。