Azure API pricing

35 models · USD per 1M tokens · verified 2026-10-23

Azure serves 35 of the models we track, 5 of them open-weights. The cheapest is Claude Haiku 5.5 at $0.10 input and $0.50 output per million tokens. For 30 of these models, Azure is the cheapest tracked host.

// models on Azure

modellabin $/1Mout $/1Mctxvs best price
Claude Haiku 5.5 BESTAnthropic$0.10$0.501Mbest
GPT-6 Luna BESTOpenAI$0.10$0.501.05Mbest
GPT-4o mini BESTOpenAI$0.15$0.60128Kbest
DeepSeek V4 FlashDeepSeek$0.19$0.511.05M+140%
GPT-5.6 Luna BESTOpenAI$0.20$1.201.05Mbest
GPT-5.4 nano BESTOpenAI$0.20$1.25400Kbest
DeepSeek V4.1 FlashDeepSeek$0.30$1.201.05M+75%
GPT-4.1 mini BESTOpenAI$0.40$1.601.05Mbest
GPT-5.4 mini BESTOpenAI$0.75$4.50400Kbest
Kimi K2.6Moonshot AI$0.95$4.00262.1Kbest
Kimi K2.7 Code HighSpeedMoonshot AI$0.95$4.00262.1K−50%
DeepSeek V4 ProDeepSeek$1.74$3.481.05M+33.8%
Grok 4.6 BESTxAI$2.00$6.00500Kbest
GPT-5.1 Codex Max BESTOpenAI$1.25$10.00400Kbest
GPT-4.1 BESTOpenAI$2.00$8.001.05Mbest
o3 BESTOpenAI$2.00$8.00200Kbest
Claude Sonnet 5 BESTAnthropic$2.00$10.001Mbest
Claude Sonnet 5.5 BESTAnthropic$2.00$10.001Mbest
GPT-6 Sol BESTOpenAI$2.00$10.001.05Mbest
GPT-6.1 Sol BESTOpenAI$2.00$10.001.05Mbest
GPT-4o BESTOpenAI$2.50$10.00128Kbest
GPT-5.6 Terra BESTOpenAI$2.00$12.001.05Mbest
GPT-5.4 BESTOpenAI$2.50$15.001.05Mbest
Claude Sonnet 4.6 BESTAnthropic$3.00$15.001Mbest
Claude Opus 5.5 BESTAnthropic$4.00$20.001Mbest
GPT-5.6 Sol BESTOpenAI$4.00$20.001.05Mbest
Claude Opus 4.7 BESTAnthropic$5.00$25.001Mbest
Claude Opus 4.8 BESTAnthropic$5.00$25.001Mbest
Claude Opus 5 BESTAnthropic$5.00$25.001Mbest
GPT-5.5 BESTOpenAI$5.00$30.001.05Mbest
Claude Fable 5 BESTAnthropic$10.00$50.001Mbest
Claude Fable 5.1 BESTAnthropic$10.00$50.001Mbest
GPT-6 Astra BESTOpenAI$10.00$50.001.05Mbest
GPT-5 Pro BESTOpenAI$15.00$120.00400Kbest
GPT-5.4 Pro BESTOpenAI$30.00$180.001.05Mbest

// how to read this page

Prices are Azure's base-tier list prices per million tokens. "vs best price" compares Azure's blended price (3 × input + output) / 4 with the lowest price we track for the same model, whether it comes from the lab's own API or another host. Open each model for its full price history and every host that serves it.