价格趋势追踪器
分析 300 个AI模型的价格分布,以及哪些服务商提供最优惠的价格。对比输入与输出成本、价格层级,并在每个预算范围内找到评分最高的模型。
价格概览
付费模型
284
免费模型
16
中位数 $/1M
$1.71
最便宜付费
Ling-2.6-flash
$0.02/1M
最贵
o1-pro
$375.00/1M
价格悬崖：免费模型最高评分达到 81。评分低于此阈值的付费模型可能不值得其成本。
价格与评分（付费模型）
LMMarketCap.com
各价格层级模型数
LMMarketCap.com
价格层级分布
| 层级 | 数量 | 平均评分 |
|---|---|---|
| Free | 16 | 46 |
| Ultra-Budget | 62 | 59 |
| Budget | 87 | 66 |
| Mid-Range | 95 | 74 |
| Premium | 20 | 74 |
| Enterprise | 20 | 84 |
完整价格表
按评分排名前 80 个模型| # | 模型 | 评分 | 平均成本 |
|---|---|---|---|
| 1 | Claude Fable 5Anthropic | 97 | $30.00 |
| 2 | Claude Fable 5 (batch)Anthropic | 97 | $15.00 |
| 3 | Claude Opus 5 (Fast)Anthropic | 95 | $30.00 |
| 4 | Claude Opus 5Anthropic | 95 | $15.00 |
| 5 | Claude Opus 4.8 (Fast)Anthropic | 95 | $30.00 |
| 6 | Claude Opus 4.8Anthropic | 95 | $15.00 |
| 7 | Claude Opus 4.7 (Fast)Anthropic | 95 | $90.00 |
| 8 | Claude Opus 4.7Anthropic | 95 | $15.00 |
| 9 | Claude Opus 4.7 (batch)Anthropic | 95 | $7.50 |
| 10 | Claude Opus 4.8 (batch)Anthropic | 95 | $7.50 |
| 11 | GPT-5.5 ProOpenAI | 93 | $105.00 |
| 12 | GPT-5.5 Pro (batch)OpenAI | 93 | $52.50 |
| 13 | GPT-5.5OpenAI | 93 | $17.50 |
| 14 | GPT-5.5 (batch)OpenAI | 93 | $8.75 |
| 15 | Gemini 3.1 Pro Preview Custom ToolsGoogle | 92 | $7.00 |
| 16 | Gemini 3.1 Pro PreviewGoogle | 92 | $7.00 |
| 17 | Gemini 3.1 Pro Preview (batch)Google | 92 | $3.50 |
| 18 | GPT-5.4 ProOpenAI | 92 | $105.00 |
| 19 | GPT-5.4 Pro (batch)OpenAI | 92 | $52.50 |
| 20 | GPT-5.4OpenAI | 92 | $8.75 |
| 21 | GPT-5.4 (batch)OpenAI | 92 | $4.38 |
| 22 | GPT-5.3-CodexOpenAI | 91 | $7.88 |
| 23 | GPT-5.2-CodexOpenAI | 91 | $7.88 |
| 24 | GPT-5.2 ChatOpenAI | 91 | $7.88 |
| 25 | GPT-5.2 ProOpenAI | 91 | $94.50 |
| 26 | GPT-5.2 Pro (batch)OpenAI | 91 | $47.25 |
| 27 | GPT-5.2OpenAI | 91 | $7.88 |
| 28 | GPT-5.2 (batch)OpenAI | 91 | $3.94 |
| 29 | Claude Opus 4.6Anthropic | 90 | $15.00 |
| 30 | Claude Opus 4.6 (batch)Anthropic | 90 | $7.50 |
| 31 | GPT-5.6 Luna ProOpenAI | 89 | $0.35 |
| 32 | GPT-5.6 Luna Pro (batch)OpenAI | 89 | $0.35 |
| 33 | GPT-5.6 LunaOpenAI | 89 | $0.35 |
| 34 | GPT-5.6 Luna (batch)OpenAI | 89 | $0.35 |
| 35 | GPT-5.6 Terra ProOpenAI | 89 | $3.50 |
| 36 | GPT-5.6 Terra Pro (batch)OpenAI | 89 | $3.50 |
| 37 | GPT-5.6 TerraOpenAI | 89 | $3.50 |
| 38 | GPT-5.6 Terra (batch)OpenAI | 89 | $3.50 |
| 39 | GPT-5.6 Sol ProOpenAI | 89 | $17.50 |
| 40 | GPT-5.6 Sol Pro (batch)OpenAI | 89 | $8.75 |
| 41 | GPT-5.6 SolOpenAI | 89 | $17.50 |
| 42 | GPT-5.6 Sol (batch)OpenAI | 89 | $8.75 |
| 43 | Grok 4.5xAI | 89 | $4.00 |
| 44 | Grok 4.3xAI | 89 | $1.88 |
| 45 | Grok 4.20xAI | 89 | $1.88 |
| 46 | GPT-5.1-Codex-MaxOpenAI | 89 | $5.63 |
| 47 | GPT-5.1OpenAI | 89 | $5.63 |
| 48 | GPT-5.1-CodexOpenAI | 89 | $5.63 |
| 49 | GPT-5 ProOpenAI | 89 | $67.50 |
| 50 | GPT-5 Pro (batch)OpenAI | 89 | $33.75 |
| 51 | GPT-5 Codex (batch)OpenAI | 89 | $2.81 |
| 52 | GPT-5OpenAI | 89 | $5.63 |
| 53 | GPT-5 (batch)OpenAI | 89 | $2.81 |
| 54 | Gemini 3 Flash PreviewGoogle | 88 | $1.75 |
| 55 | Gemini 3 Flash Preview (batch)Google | 88 | $0.88 |
| 56 | Grok 4.20 Multi-AgentxAI | 88 | $1.88 |
| 57 | GPT-5.1 (batch)OpenAI | 88 | $2.81 |
| 58 | DeepSeek V4 ProDeepSeek | 87 | $0.95 |
| 59 | o3 ProOpenAI | 87 | $50.00 |
| 60 | o3 Pro (batch)OpenAI | 87 | $25.00 |
| 61 | o3OpenAI | 87 | $5.00 |
| 62 | o3 (batch)OpenAI | 87 | $2.50 |
| 63 | Claude Sonnet 5Anthropic | 85 | $6.00 |
| 64 | Claude Sonnet 4.6Anthropic | 85 | $9.00 |
| 65 | Claude Sonnet 4.6 (batch)Anthropic | 85 | $4.50 |
| 66 | Claude Opus 4.5Anthropic | 85 | $15.00 |
| 67 | Claude Opus 4.5 (batch)Anthropic | 85 | $7.50 |
| 68 | Gemini 2.5 ProGoogle | 84 | $5.63 |
| 69 | Gemini 2.5 Pro (batch)Google | 84 | $2.81 |
| 70 | Gemini 2.5 Pro Preview 06-05Google | 84 | $5.63 |
| 71 | Gemini 2.5 Pro Preview 05-06Google | 84 | $5.63 |
| 72 | GLM 5.2Zhipu AI | 83 | $1.59 |
| 73 | GLM 5.2 (batch)Zhipu AI | 83 | $1.45 |
| 74 | Claude Sonnet 4.5Anthropic | 82 | $9.00 |
| 75 | Claude Sonnet 4.5 (batch)Anthropic | 82 | $4.50 |
| 76 | GLM 5.1Zhipu AI | 82 | $2.90 |
| 77 | GLM 5 TurboZhipu AI | 82 | $2.60 |
| 78 | GLM 5Zhipu AI | 82 | $1.75 |
| 79 | Claude Opus 4.1Anthropic | 82 | $45.00 |
| 80 | Claude Opus 4Anthropic | 82 | $45.00 |
还有 220 个模型在追踪中。
输入与输出价格分析
I/O比 = 输出价格 / 输入价格。3.0x 的比率意味着输出令牌的成本是输入令牌的3倍。
最高输入输出比
输出相对输入最贵| 模型 | $/M In | $/M Out | 比率 |
|---|---|---|---|
| Qwen3 Next 80B A3B Instruct | $0.09 | $1.10 | 12.2x |
| Qwen3 30B A3B Thinking 2507 | $0.20 | $2.40 | 12.0x |
| Perceptron Mk1 | $0.15 | $1.50 | 10.0x |
| Palmyra X5 | $0.60 | $6.00 | 10.0x |
| Qwen3 VL 235B A22B Thinking | $0.40 | $4.00 | 10.0x |
| Qwen3 235B A22B Thinking 2507 | $0.23 | $2.30 | 10.0x |
| Qwen3 VL 235B A22B Instruct | $0.21 | $1.90 | 9.0x |
| Gemini 3.5 Flash Lite | $0.30 | $2.50 | 8.3x |
| Gemini 3.5 Flash Lite (batch) | $0.15 | $1.25 | 8.3x |
| Ring-2.6-1T | $0.07 | $0.63 | 8.3x |
最低输入输出比
价格最均衡| 模型 | $/M In | $/M Out | 比率 |
|---|---|---|---|
| Reka Edge | $0.10 | $0.10 | 1.0x |
| R1 Distill Llama 70B | $0.80 | $0.80 | 1.0x |
| Llama 3.1 70B Instruct | $0.40 | $0.40 | 1.0x |
| Gemma 2 27B | $0.65 | $0.65 | 1.0x |
| DeepSeek V3.2 | $0.27 | $0.40 | 1.5x |
| Qwen3.5-9B | $0.10 | $0.15 | 1.5x |
| DeepSeek V3.2 Exp | $0.27 | $0.41 | 1.5x |
| Llama 3.1 8B Instruct | $0.05 | $0.08 | 1.6x |
| Laguna S 2.1 | $0.09 | $0.18 | 2.0x |
| Aion-3.0-Mini | $0.70 | $1.40 | 2.0x |
输入/输出价格相同的模型 (4)
各价位最佳性价比
在每个平均成本阈值内评分最高的模型。
| 最高 $/1M | 模型 | 评分 | 实际成本 |
|---|---|---|---|
| $0.10 | Gemma 4 31B (free) | 81 | 免费 |
| $0.50 | GPT-5.6 Luna Pro | 89 | $0.35/1M |
| $1.00 | GPT-5.6 Luna Pro | 89 | $0.35/1M |
| $2.00 | GPT-5.6 Luna Pro | 89 | $0.35/1M |
| $5.00 | Gemini 3.1 Pro Preview (batch) | 92 | $3.50/1M |
| $10.00 | Claude Opus 4.7 (batch) | 95 | $7.50/1M |
| $20.00 | Claude Fable 5 (batch) | 97 | $15.00/1M |
| $50.00 | Claude Fable 5 | 97 | $30.00/1M |
服务商定价策略
拥有2个以上模型的服务商，按平均成本排序| 提供商 | 模型 | 平均成本 |
|---|---|---|
| TII | 3 | Free |
| poolside | 4 | $0.06 |
| inclusionai | 5 | $0.15 |
| Upstage | 2 | $0.23 |
| Tencent | 2 | $0.23 |
| Meta | 5 | $0.26 |
| Xiaomi | 2 | $0.43 |
| StepFun | 2 | $0.44 |
| NVIDIA | 7 | $0.48 |
| ByteDance | 4 | $0.67 |
| DeepSeek | 12 | $0.69 |
| MiniMax | 8 | $0.74 |
| Kuaishou | 3 | $0.99 |
| Alibaba | 31 | $1.21 |
| Zhipu AI | 13 | $1.47 |
| thinkingmachines | 3 | $1.53 |
| meta | 3 | $2.14 |
| xAI | 5 | $2.23 |
| aion-labs | 3 | $2.25 |
| 27 | $2.26 | |
| Moonshot AI | 8 | $2.63 |
| Mistral AI | 6 | $2.98 |
| Cursor | 2 | $3.00 |
| Cohere | 4 | $3.22 |
| 2 | $5.75 | |
| sakana | 2 | $9.99 |
| ~openai | 2 | $10.06 |
| ~anthropic | 4 | $13.50 |
| Anthropic | 28 | $16.79 |
| OpenAI | 85 | $19.06 |
Frequently Asked Questions
We categorize models into tiers based on average cost per million tokens: Free (zero cost), Budget (under $1), Mid ($1-$10), Premium ($10-$50), and Enterprise (over $50). This tiering helps developers quickly find models that match their budget constraints.
The I/O ratio compares the cost of output tokens to input tokens. A ratio of 3x means output costs three times more than input. This matters because output-heavy workloads (like content generation) will cost more with high-ratio models, while input-heavy workloads (like analysis) are less affected.
Use the "Best Value at Each Price Point" table to find the highest-scoring model within your budget. Set your maximum cost threshold, and the table shows which model delivers the best quality at or below that price. You can also check the scatter chart to visually identify models with the best score-to-cost ratio.