AI API Gateway Rates
The AI API Gateway service is available to all Stanford faculty, staff, and students with a valid PTA.
| Language Model | Input cost ¹ (Price per 1M tokens) | Output cost ¹ (Price per 1M tokens) |
|---|---|---|
| GPT 4o mini | 50% off standard OpenAI prices | 50% off standard OpenAI prices |
| GPT 4o | 50% off standard OpenAI prices | 50% off standard OpenAI prices |
| GPT o1 | 50% off standard OpenAI prices | 50% off standard OpenAI prices |
| GPT o3 mini | 50% off standard OpenAI prices | 50% off standard OpenAI prices |
| GPT 4.1 | 50% off standard OpenAI prices | 50% off standard OpenAI prices |
| OpenAI Text-embedding-ada-002 | 50% off standard OpenAI prices | 50% off standard OpenAI prices |
| GPT 5 | 50% off standard OpenAI prices | 50% off standard OpenAI prices |
| GPT 5 mini | 50% off standard OpenAI prices | 50% off standard OpenAI prices |
| GPT 5.2 | 50% off standard OpenAI prices | 50% off standard OpenAI prices |
| GPT 5.4 | 50% off standard OpenAI prices | 50% off standard OpenAI prices |
| GPT 5.5 | 50% off standard OpenAI prices | 50% off standard OpenAI prices |
| GPT 5.6 Luna | 50% off standard OpenAI prices | 50% off standard OpenAI prices |
| GPT 5.6 Terra | 50% off standard OpenAI prices | 50% off standard OpenAI prices |
| GPT 5.6 Sol | 50% off standard OpenAI prices | 50% off standard OpenAI prices |
| Gemini 2.5 Pro | $ 2.50 | $ 15.00 |
| Gemini 2.5 Flash | $ 0.30 | $ 2.50 |
| Gemini 2.5 Flash Lite | $ 0.10 | $ 0.40 |
| Claude 4.5 Haiku | 50% off standard Anthropic prices | 50% off standard Anthropic prices |
| Claude 3.5 Sonnet | 50% off standard Anthropic prices | 50% off standard Anthropic prices |
| Claude 3.7 Sonnet | 50% off standard Anthropic prices | 50% off standard Anthropic prices |
| Claude 4.5 Sonnet | 50% off standard Anthropic prices | 50% off standard Anthropic prices |
| Claude 4.6 Sonnet | 50% off standard Anthropic prices | 50% off standard Anthropic prices |
| Claude 5 Sonnet | 50% off standard Anthropic prices | 50% off standard Anthropic prices |
| Claude 4.6 Opus | 50% off standard Anthropic prices | 50% off standard Anthropic prices |
| Claude 4.7 Opus | 50% off standard Anthropic prices | 50% off standard Anthropic prices |
| Claude 4.8 Opus | 50% off standard Anthropic prices | 50% off standard Anthropic prices |
| Claude 5 Opus | 50% off standard Anthropic prices | 50% off standard Anthropic prices |
| Llama 4 | $ 0.35 | $ 1.15 |
| Llama 3.2 | No published price. ² | No published price. ² |
| DeepSeek R1 | No published price. ² | No published price. ² |
| DeepSeek V3 0324 | No published price. ² | No published price. ² |
| DeepSeek V3.2 | No published price. ² | No published price. ² |
| Image Model | Image generation cost ¹ (Price per image) | Input cost ¹ (Price per 1M tokens) | Output cost ¹ (Price per 1M tokens) |
|---|---|---|---|
| Google Imagen 3 | $ 0.04 | N/A | N/A |
| Google Imagen 4 | $ 0.06 | N/A | N/A |
| Gemini Flash Image 2.5 | N/A | $ 0.30 | $ 30.00 ³ |
¹ All prices are published by the vendors and have Stanford discounts already applied if applicable.
² Models available in preview may be subject to additional vendor rate limits and the charges could change at any time.
³ A 1024x1024 image consumes 1290 tokens. Per image token count varies by image resolution. For more information on how to calculate tokens, you can refer to Google's documentation.
⁴ Price not discounted off standard prices.
Pricing of any and all models may change at any time, without warning. We will do our best to update this page as soon as vendor updates are made public. Continued usage of models will be subject to the new price published by the vendors at the time of the vendor's change.
For more information about the AI API Gateway, visit the AI API Gateway service page.
