Accessing Generative AI APIs at ASU
1. Academic Research
ASU has an enterprise OpenAI platform account for academic use.
If you are conducting research as part of ASU faculty, staff, or a sponsored student project:
- Kindly visit: https://rto.asu.edu/ai-support-for-research/ and submit the form.
KE Contact: Gil Speyer (Knowledge Enterprise)
2. Enterprise Development (University-Supported Units)
If you’re part of a development team supporting official ASU initiatives (e.g., EdPlus, Learning Enterprise, academic platforms):
- Access: CreateAI APIs give you access to over 50+ models, including OpenAI, Anthropic, Mistral, Cohere, and more with enterprise-grade security controls aligned with ASU data governance.
Please request your API key directly within CreateAI.
Go to your project → Profile → API Keys → Request API Key, then submit the form.
Your API key will typically be available within 24–48 hours (often sooner). If it’s urgent, feel free to message our team.
Learn more: API Access Guide
- Click API Access For Generative AI at ASU for more details on the CreateAI API Access and Usage Policy
3. Individual Development (Non-Research, Not Official ASU Project)
If your project is more exploratory or doesn’t fall under research or enterprise dev:
- Option 1: Work with your unit’s IT and procurement team to purchase direct API access from OpenAI or another vendor.
- Option 2: Explore AWS Bedrock, which provides access to multiple foundation models (e.g., Anthropic, Cohere).
Option 3: Use Microsoft Azure with OpenAI credits, if already available through your unit
🔐 Note: These options will not fall under ASU’s enterprise contract. You will be required to complete a VITRA security review to ensure proper data protections are in place.
Summary Table
| Use Case | Contact | Access Type |
|---|---|---|
Academic Research | Gil Speyer ([email protected]) | OpenAI API via ASU research license |
Enterprise Dev (EdPlus, Learning Enterprise, etc.) | Roger Kohler ([email protected]) | CreateAI APIs (50+ models) |
Individual / Prototyping | Your IT/Procurement Team | Direct OpenAI, AWS Bedrock, Azure APIs (requires VITRA) |
Keep Reading
Breakdown of RAG Model Parameters, Settings and Their Impact
Kofi Wood
Retrieval-Augmented Generation (RAG) is an advanced approach in natural language processing that integrates information retrieval and generative language modeling. Unlike traditional language models that generate responses solely based on their pre-trained knowledge, RAG combines retrieval mechanisms with generative models to enhance the relevance and accuracy of its responses. This hybrid framework works by first retrieving relevant documents or information from a predefined knowledge base (e.g., databases, documents, or PDFs) and then using a generative model (such as a transformer-based model) to synthesize a response that incorporates the retrieved context.