Empower enterprises to access top-tier models, deploy with confidence and optimize for performance
浏览定价选项
应用筛选器来根据你的需求自定义定价选项。
价格仅是估算值,不应用作实际报价单。实际定价可能因与 Microsoft 签订的协议类型、购买日期和货币汇率而异。价格基于美元计算,并使用在上一个月末最后一个工作日之前的两个工作日内捕获的伦敦收盘即期汇率进行转换。如果本月底之前的两个工作日在主要市场中属于银行假日,则汇率确定日通常是紧接在此两个工作日之后的一天。此汇率适用于下个月的所有交易。登录到 Azure 定价计算器查看基于你与 Microsoft 的当前计划/产品/服务的定价。有关定价或请求报价的详细信息,请联系 Azure 销售专家。请参阅经常询问有关 Azure 定价的问题。
美国政府实体有资格从授权解决方案提供商处购买 Azure 政府服务(无需预付定金),或者可直接通过即用即付在线订阅购买。
重要提示 - 该价格 (R$) 只是一个参考;这是一项国际交易,最终价格受汇率和所含 IOF 税的影响。不会发布 eNF。
美国政府实体有资格从授权解决方案提供商处购买 Azure 政府服务(无需预付定金),或者可直接通过即用即付在线订阅购买。
重要提示 - 该价格 (R$) 只是一个参考;这是一项国际交易,最终价格受汇率和所含 IOF 税的影响。不会发布 eNF。
Serverless Deployment
Azure AI Foundry Models continues to grow with cutting-edge additions—including models from OpenAI, DeepSeek, xAI’s Grok, Meta’s Llama, Mistral AI, FLUX by Black Forest Labs, and more. These models are fully hosted and managed on Azure, and available through both pay-as-you-go and provisioned throughput options. With Provisioned Throughput, you can even flex capacity across multiple models, such as OpenAI and DeepSeek, for greater efficiency and control.
AOAI
Open AI Pricing Page: Azure OpenAI Service - Pricing | Microsoft Azure
Phi
语言模型
模型 | 上下文 | 输入(每 1,000 个令牌) | 输出(每 1,000 个令牌) |
---|---|---|---|
Phi-3-mini-4k-instruct | 4K | $- | $- |
Phi-3-mini-128k-instruct | 128K | $- | $- |
Phi-3.5-mini-instruct | 128K | $- | $- |
Phi-3-small-8k-instruct | 8K | $- | $- |
Phi-3-small-128k-instruct | 128K | $- | $- |
Phi-3-medium-4k-instruct | 4K | $- | $- |
Phi-3-medium-128k-instruct | 128K | $- | $- |
Phi-3.5-MoE-instruct | 128K | $- | $- |
Phi-4 | 128K | $- | $- |
Phi-4-mini | 128K | $- | $- |
Phi-4-multimodal, text and image | 128K | $- | $- |
Phi-4-multimodal, audio | 128K | $- | $- |
Phi-4-mini-reasoning | 128K | $- | $- |
Phi-4-reasoning | 32K | $- | $- |
Phi-4-reasoning-plus | 32K | $- | $- |
微调模型
模型 | 每 1,000 个令牌的训练 | 每小时托管数 | 每 1,000 个令牌的输入使用情况 | 每 1,000 个令牌的输出使用情况 |
---|---|---|---|---|
Phi-3-mini-4k-instruct | $- | $- | $- | $- |
Phi-3-mini-128k-instruct | $- | $- | $- | $- |
Phi-3.5-mini-instruct | $- | $- | $- | $- |
Phi-3-medium-4k-instruct | $- | $- | $- | $- |
Phi-3-medium-128k-instruct | $- | $- | $- | $- |
Phi-3.5-MoE-instruct | $- | $- | $- | $- |
Phi-4 | $- | $- | $- | $- |
Phi-4-mini | $- | $- | $- | $- |
DeepSeek is a family of large language models (LLMs) developed by the DeepSeek AI research team. These models are designed to perform a wide range of natural language processing (NLP) and reasoning tasks, with a focus on open-source accessibility, performance optimization, and multi-modal capabilities.
语言模型
模型 | 输入(每 1,000 个令牌) | 输出(每 1,000 个令牌) |
---|---|---|
DeepSeek R1 Global | $- | $- |
DeepSeek R1 DataZone | $- | $- |
DeepSeek R1 Regional | $- | $- |
DeepSeek V3 Global | $- | $- |
DeepSeek V3 Regional | $- | $- |
DeepSeek V3 0324 Global | $- | $- |
DeepSeek V3 0324 Regional | $- | $- |
Provisioned
模型 | Min PTU | PTU Hourly Pricing | PTU Monthly Reservation Pricing |
---|---|---|---|
DeepSeek R1 Global | 100 | $- | $- |
DeepSeek V3 Global | 100 | $- | $- |
DeepSeek V3 0324 Global | 100 | $- | $- |
模型 | 输入(每 1,000 个令牌) | 输出(每 1,000 个令牌) |
---|---|---|
MAI-DS-R1 Global | $- | $- |
MAI-DS-R1 Regional | $- | $- |
Provisioned
模型 | Min PTU | PTU Hourly Pricing | PTU Monthly Reservation Pricing |
---|---|---|---|
MAI-DS-R1 Global | 100 | $- | $- |
模型 | 输入(每 1,000 个令牌) | 输出(每 1,000 个令牌) |
---|---|---|
Grok-3 Global | $- | $- |
Grok-3 DataZone | $- | $- |
Grok-3 Regional | $- | $- |
Grok-3 Mini Global | $- | $- |
Grok-3 Mini DataZone | $- | $- |
Grok-3 Mini Regional | $- | $- |
模型 | 输入(每 1,000 个令牌) | 输出(每 1,000 个令牌) |
---|---|---|
Llama 3.3 70B Datazone | $- | $- |
Llama 3.3 70B Global | $- | $- |
Llama 3.3 70B Regional | $- | $- |
Llama4 Maverick 17B Datazone | $- | $- |
Llama4 Mavrick 17B Global | $- | $- |
Llama4 Maverick 17B Regional | $- | $- |
Azure 定价和购买选项
其他资源
Azure AI Foundry Models
详细了解 Azure AI Foundry Models 特性和功能。
定价计算器
估计每月使用任何 Azure 产品组合应产生的费用。
文档
查看技术教程、视频和更多 Azure AI Foundry Models 资源。
与销售专家交谈,演练 Azure 定价情况。了解你的云解决方案的定价。
获取免费云服务和价值 $200 的赠金来探索 Azure 30 天。