You can use the Manage Keys page at https://runchat.app/dashboard/keys to add and encrypt your API keys for other services to use them securely in Runchat.
Runchat will automatically extend the functionality of some nodes when you add supported services. These include using Fal to add models to the Image node, and OpenRouter, Groq and Cerebras to add models to the Prompt node.
You can use Runchat entirely for free by linking an OpenRouter account. OpenRouter serves models from most of the major research labs, including experimental and free models with rate limits. To connect OpenRouter, visit https://openrouter.ai and create an account. Then from the keys dashboard at https://runchat.app/dashboard/keys click Authorize OpenRouter. This will open a popup where you will be prompted to sign in to your OpenRouter account and then Authorize Runchat. Enter a max spend cap if you want (if you don’t have any open router credits you don’t need to) then click Authorize the return to the Runchat Dashboard. You can now install models from OpenRouter on the models page and these will appear as options in the Prompt node settings.
<aside> 🎬
Watch a quick video to see how to integrate with fal to generate better images:
https://www.youtube.com/watch?v=wj4l8CwuIik
</aside>
Fal is a fast inference provider for generative AI models. They host models to generate and edit images, video, audio and text. Runchat uses Fal as the default provider for image generation in the Image Node. When you connect Fal, Runchat will automatically add options to the Image Node settings that will allow you to choose from fast and low-cost models (Stable Diffusion Turbo), models with a good tradeoff of price and quality (Flux schnell and Flux Dev) and high quality models (Flux Pro and Recraft). Recraft can be used to generate vector graphics as well as raster images.
<aside> 💡
For more information on the Image Node check the Guide.
</aside>
To connect Fal, head to fal.ai and create an account. Then visit https://fal.ai/dashboard/keys and click Add Key. Leave the scope as “API”, enter a name to remember the key by (e.g. runchat) and click Create Key. Then copy the generated key, paste it into the Fal section on https://runchat.app/dashboard/keys and click Add Fal Key to encrypt it and add it to your Runchat account. Runchat will automatically detect this key and enable options to use high quality models in the Image node in the editor.
Groq and Cerebras are custom chip manufacturers that offer cloud APIs for running very fast inference of language models. These chips can generate thousands of tokens per second, resulting in near instant responses from the Prompt node. This in turn allows you to chain many prompt nodes together without having to sit around waiting for your Runchat to update every time you make a change.
<aside> 💡
For more information on the Prompt Node check the Guide.
</aside>
To connect Groq, head to https://console.groq.com/ and sign up. Then visit https://console.groq.com/keys and click Create API Key. Enter a name (e.g. runchat) and click Submit. Then copy the generated key, paste it into the Groq section on https://runchat.app/dashboard/keys and click Add Groq Key to encrypt it and add it to your Runchat account. Runchat will automatically detect this key and display Groq models in the Prompt node settings. Groq runs the latest Llama models with extremely fast inference times.
Cerebras is another very fast inference provider for language models. To connect Cerebras, visit https://cloud.cerebras.ai/ and sign up. Then click on API Keys from the sidebar, and click Generate API Key. Give you key a name. Then copy the generated key, paste it into the Cerebras section on https://runchat.app/dashboard/keys and click Add Cerebras Key to encrypt it and add it to your Runchat account. Runchat will automatically detect this key and display Cerebras models in the Prompt node settings.
Working with Google Sheets requires authorizing Runchat to access to your Google Sheets account via OAuth. Click Authorize Google Sheets from the https://runchat.app/dashboard/keys page to display the OAuth consent screen. Sign into your Google account and then follow the prompts to Authorize Runchat and return to the dashboard. You can now work with the Google Sheets api in the API node using the google_sheets variable to reference your encrypted access token.