Gemini Pro API is now live

Google has made Gemini Pro available in API. Gemini Pro is the 2nd-best model from its new family of AI models Gemini. Gemini Pro is currently free for developers and enterprises to start building with.

What’s going on here?

Gemini Pro is now accessible via API for app development.

What does this mean?

Here’s what you have access to right now (not gonna bother with what’s coming up).

• 32K context window for text and vision endpoint for multimodal use.

• Free of cost until general availability next year. Limited at 60 queries per minute.

• Features: function calling, embeddings, semantic retrieval and custom knowledge grounding, and chat functionality.

• SDKs for Python, Kotlin, Node.js, Swift, and Javascript.

• Can be used via Google AI Studio and Vertex AI.

The pricing for Gemini Pro (after general availability) is almost similar to the comparable model GPT-3.5-turbo.

For enterprise users, Google Cloud has also added a ton of things like Duet AI for code, AI Hypercomputer and other hardware upgrades.

Why should I care?

Google claims it outperforms similarly sized models, offers multilingual support across 180+ countries. But as everyone predicted, giving devs access a week after the launch is hurting Google. I haven’t seen any dev playing with it on X. (Mistral is where the attention is at)

But if you want to try the models out, testing it now is the best option. The free options are generous and there’s generally alpha in looking where others aren’t.