LLM from Google can finally do something
Now Gemini 2.0 Flash as an experiment is free to use through Google AI Studio and API. It has a large context window - 2 million tokens (this is for input). 😲
Models from Google can be connected via API in Cursor settings (tab "Models"), but auto-patching from chat disappears and requires the Pro version, in my opinion, this is stinginess.
And in this video
about how to use Gemini 2.0 Flash using the tools Aider (for terminal) and Cline for VSC. 💻
- demonstrates how to install and configure these tools. Talks about limitations through the API. ⚙️
- notes that Gemini 2.0 Flash works very fast and has excellent function calling capabilities. ⚡️
On the Aider leaderboard, it is of course behind the top models, but you have to pay for them, and this is a Flash series model, i.e., simplified.