⚡ An all-in-one managed backend for AI apps. Build AI apps from the frontend, very fast. 🪶
Why use Polyfire?
- Just code from the frontend, no backend needed
- If you already have backend, less code to write
- Most backend services you'd need in a couple lines of code
We manage your AI backend so you don't have to.
import { useState, useEffect } from "react";
import { createRoot } from "react-dom/client";
import { TextGenerated, Login } from "polyfire-js/components";
function App() {
return (
<Login>
<h2>Here's a little auto-generated haiku for you:</h2>
<TextGenerated prompt="Generate a hello world haiku" />
</Login>
);
}
document.body.innerHTML = '<div id="app"></div>';
const root = createRoot(document.getElementById("app"));
root.render(
<PolyfireProvider project="your_project_id">
<App />
</PolyfireProvider>,
);Don't forget to change the your_project_id by your project ID you will have got on https://beta.polyfire.com
<script src="https://github.com/polyfire-ai/polyfire-js/releases/download/0.2.7/polyfire-min-0.2.7.js"></script>
<script>
(async () => {
const polyfire = window.PolyfireClientBuilder({ project: "your_project_id" });
const isAuthenticated = await polyfire.auth.init();
if (!isAuthenticated) {
await polyfire.auth.login("github");
}
const helloWorld = await polyfire.models.generate("Write me a hello world haiku");
console.log(helloWorld);
})();
</script>Don't forget to change the your_project_id by your project ID you will have got on https://beta.polyfire.com
- Website: polyfire.com
- Dashboard: beta.polyfire.com
- Discord: polyfire.com/discord
- Javascript SDK: github.com/polyfact/polyfire-js
- Open Source API (your managed backend!): github.com/polyfact/polyfire-api
We're open source! Make a good PR to the JS SDK or the API and we'll merge it.
This SDK provides functionalities for text generation, transcription, memory management, image generation, and data loading. Below is a guide on how to use these features.
To install the SDK, use the following command:
npm install polyfire-jsTo use the SDK, import the necessary functions and classes:
import {
generate,
generateWithType,
transcribe,
Chat,
createMemory,
updateMemory,
getAllMemories,
generateImage,
TextFileLoader,
StringLoader,
AudioLoader,
kv,
usage,
t
} from "polyfire-js";
import PolyfireClientBuilder from "polyfire-js";- generate: Generate text based on input.
- generateWithType: Generate text with probabilistic types.
const options: GenerationOptions = { /* Generation options */ };
const result = await generate("Your input text", options);- transcribe: Transcribe audio to text.
const transcription = await transcribe(audioFile);- Chat: Class for handling chat functionalities.
const chat = new Chat();
chat.sendMessage("Hello!");- createMemory: Create embeddings for memory.
- updateMemory: Update existing embeddings.
- getAllMemories: Retrieve all embeddings.
const memory = createMemory(data);
const updatedMemory = updateMemory(memoryId, newData);
const allMemories = getAllMemories();- generateImage: Generate images based on input.
const image = await generateImage("A beautiful sunset");- TextFileLoader: Load text files.
- StringLoader: Load strings.
- AudioLoader: Load audio files.
const textLoader = new TextFileLoader(filePath);
const stringLoader = new StringLoader(stringData);
const audioLoader = new AudioLoader(audioFile);- kv.get: Retrieve a value by key.
- kv.set: Store a value by key.
- kv.del: Delete a value by key.
- kv.all: Retrieve all key-value pairs.
const value = await kv.get("key");
await kv.set("key", "value");
await kv.del("key");
const allValues = await kv.all();- usage: Track usage of the SDK.
const usageData = await usage();- t: Type validation using
polyfact-io-ts.
const isValid = t.validate(data, schema);- PolyfireClientBuilder: Build the client.
const client = new PolyfireClientBuilder({ apiKey: "your-api-key" });Here is a complete example of using the SDK:
import {
generate,
transcribe,
createMemory,
generateImage,
kv,
usage
} from "polyfire-js";
import PolyfireClientBuilder from "polyfire-js";
async function main() {
// Text generation
const text = await generate("Hello, world!");
// Transcription
const transcription = await transcribe("path/to/audio/file");
// Memory management
const memory = createMemory("Some data");
// Image generation
const image = await generateImage("A beautiful sunset");
// Key-Value operations
await kv.set("key", "value");
const value = await kv.get("key");
// Usage tracking
const usageData = await usage();
// Client builder
const client = new PolyfireClientBuilder({ apiKey: "your-api-key" });
}
main();