A guide to the models available on campusGenAI, including how they are accessed, what they are commonly used for, and their knowledge cutoff.
OpenAI
| Model name | Description | Knowledge cutoff | Provider | Cost ($–$$$) |
|---|---|---|---|---|
| GPT-5-mini | A faster, more cost-efficient version of GPT-5. Optimized for well-designed tasks and precise prompts. Trained for code generation, instruction following, and long text and tool calling. Includes web search and image generation capabilities. | May 2024 | Azure | $ |
| GPT-5.2 | A powerful next-generation model that delivers the deep reasoning and expanded context handling necessary for building sophisticated agents. Includes web search and image generation capabilities. | August 2025 | Azure | $$ |
| GPT-5 | A powerful next-generation model that offers improved reasoning, context depth, and creativity. Trained for code generation, instruction following, and long text and tool calling. Includes web search and image generation capabilities. | September 2024 | Azure | $$ |
| GPT-5 Codex | An advanced code generation model built on GPT-5 that offers strong reasoning, debugging, and language support. Proficient in coding, technical explanations, and software development. | August 2025 | Azure | $$ |
| GPT-4o | A fast, versatile model for advanced tasks. Provides improved reasoning, multimodal capabilities, and coding assistance. | October 2023 | Azure | $$ |
Anthropic
| Model name | Description | Knowledge cutoff | Provider | Cost ($–$$$) |
|---|---|---|---|---|
| Claude 4.6 Sonnet | A hybrid-reasoning model for agentic tasks including coding, long-form reasoning, and tool and agent planning. Includes web search and image generation capabilities. | August 2025 | AWS Bedrock | $$$ |
| Claude 4.6 Opus | A flagship model for agentic coding, tool and computer use, long-context reasoning, and enterprise workflows. Includes web search and image generation capabilities. | May 2025 | AWS Bedrock | $$$ |
| Claude 4.5 Sonnet | A smart model for quick, step-by-step tasks. Designed for coding, using computers, reasoning, and mathematics. Ideal for building complex agents. Includes web search and image generation capabilities. | January 2025 | AWS Bedrock | $$$ |
| Claude 4.5 Haiku | A lightweight, fast model for efficient chatting and analysis. Balances coding, reasoning, and analysis, with low costs and high speeds. Includes web search and image generation capabilities. | February 2025 | AWS Bedrock | $ |
DeepSeek
| Model name | Description | Knowledge cutoff | Provider | Cost ($–$$$) |
|---|---|---|---|---|
| DeepSeek-R1 | A strong model for deep reasoning and code. Offers advanced deep reasoning and comprehension capabilities and is designed for efficient natural language processing. | July 2024 | Azure | $$ |
Meta
| Model name | Description | Knowledge cutoff | Provider | Cost ($–$$$) |
|---|---|---|---|---|
| LLaMA 3 70B | A top model for instruction following and nuanced reasoning and accuracy. | December 2023 | AWS Bedrock | $$ |
Mistral
| Model name | Description | Knowledge cutoff | Provider | Cost ($–$$$) |
|---|---|---|---|---|
| Mistral Large | A model suitable for complex tasks that require substantial deep reasoning. | October 2023 | AWS Bedrock | $ |
xAI
| Model name | Description | Knowledge cutoff | Provider | Cost ($–$$$) |
|---|---|---|---|---|
| Grok 3 | A fast reasoning model for diverse tasks. Focuses on real-time conversational capabilities, current event awareness, and witty engaging dialogue. Ideal for social media interactions, trending topic discussions, and conversational tasks needing up-to-date context. | November 2024 | Azure | $$ |
