A few call outs, fresh from our resident agent builders:
- 🐝 It can have about double the input context and about 6x the output context
- 🤖 From our testing, its agent behaviour seems a bit different to GPT-4o. It wants to use tools more than 4o did
- ⚡According to OpenAI's metrics, it is a more powerful model (including better at tool calling) ⚡
- 💰It’s ~ 6x more expensive than gpt4o
Keen to power your agents with o1? 👇
To use it in Relevance, you’ll need to bring-your-own OpenAI API key (for now!)