I prefer using Claude 3.5 Sonnet for writing complex JavaScript/Playwright logic, but I find GPT-4o is much better at the Vision Agent tasks.
Does RTILA X lock you into a single provider, or can I use OpenRouter to route different tasks to different models?
You can absolutely use OpenRouter, and it’s our recommended setup for cloud models.
In your Preferences, just select OpenRouter as your active provider and drop in your API key. Our context builder automatically formats all payloads - including complex multimodal image attachments for the Vision Agent - to be perfectly compatible with the OpenAI standard that OpenRouter expects.
You can easily swap the active model name in the settings depending on your current task. Use Claude 3.5 for heavy logic generation, and then swap to a vision-heavy model when you need to parse screenshots.
Hello, I attempted to connect RTILA X with google API key but it keeps telling it fails to connect. I tried different configurations, different models, with all the same issue. fail to connect?
Hi, RTILA X doesn’t accept direct Google or Gemini keys. To fix the connection error, you need to register at openrouter.ai, generate an API key from your account there, and put that into RTILA X instead. You will still be able to use Gemini models through OpenRouter.
Do they have the option to use the free option to try it out?