Azure OpenAI pricing
Verified Source Azure OpenAI (Foundry Models)14 current modelsUSD, standard tier, before tax
GPT-5 on Azure OpenAI Global Standard costs $1.25 per 1M input tokens and $10.00 per 1M output tokens, matching OpenAI's direct price of $1.25 / $10.00. GPT-5 mini is $0.25 / $2.00. Cached input on GPT-5 is $0.125 per 1M.
Price table
| Model | Input | Output | Cached input | Batch | Verified |
|---|---|---|---|---|---|
GPT-5azure-openai/gpt-5 | $1.25 | $10.00 | $0.125 | 50% | |
GPT-5 miniazure-openai/gpt-5-mini | $0.25 | $2.00 | $0.025 | 50% | |
GPT-5 nanoazure-openai/gpt-5-nano | $0.05 | $0.40 | $0.005 | 50% | |
GPT-5-Codexazure-openai/gpt-5-codex | $1.25 | $10.00 | $0.125 | 50% | |
GPT-5 proazure-openai/gpt-5-pro | $15.00 | $120.00 | n/a | 50% | |
GPT-4.1azure-openai/gpt-4.1 | $2.00 | $8.00 | $0.50 | 50% | |
GPT-4.1 miniazure-openai/gpt-4.1-mini | $0.40 | $1.60 | $0.10 | 50% | |
GPT-4.1 nanoazure-openai/gpt-4.1-nano | $0.10 | $0.40 | $0.025 | 50% | |
GPT-4o (2024-11-20)azure-openai/gpt-4o | $2.50 | $10.00 | $1.25 | 50% | |
GPT-4o miniazure-openai/gpt-4o-mini | $0.15 | $0.60 | $0.075 | 50% | |
o3azure-openai/o3 | $2.00 | $8.00 | $0.50 | 50% | |
o4-miniazure-openai/o4-mini | $1.10 | $4.40 | $0.275 | 50% | |
o3-proazure-openai/o3-pro | $20.00 | $80.00 | n/a | 50% | |
o1azure-openai/o1 | $15.00 | $60.00 | $7.50 | 50% |
Worked example
1,000 requests per day, each with 3,000 input tokens and 600 output tokens, on GPT-5:
- 3,000 input tokens at $1.25 per 1M costs $0.003750 per request
- 600 output tokens at $10.00 per 1M costs $0.006000 per request
- $0.009750 per request, which is $9.75 per day, $292.50 per month over 30 days and $3,558.75 per year.
- Same workload through the Batch API: $0.004875 per request, $146.25 per month.
Questions
Is Azure OpenAI more expensive than OpenAI directly?
For Global Standard deployments the per-token list prices match OpenAI's own prices for the same model. Data Zone and Regional deployments, Provisioned Throughput Units and fine-tuned deployments are priced differently and are not on this page.
Where do these numbers come from?
They are read from the Azure Retail Prices API for the eastus2 region and the Global deployment type, cross-checked against the Azure OpenAI pricing page. The API returns per 1,000 token units; SpendLint stores them per 1M.
Does Azure offer a batch discount?
Yes. Global Batch deployments are priced at 50% of the Global Standard rate for supported models.
Retired models
Sources and verification
- Azure OpenAI (Foundry Models): https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/