docs: add ZeroLimitAI custom endpoint provider

components/icons/providers/index.ts

@@ -50,3 +50,4 @@ export { default as VLLMIcon } from './vllm'
 export { default as VultrIcon } from './vultr'
 export { default as WolframIcon } from './wolfram'
 export { default as XAIIcon } from './xai'
+export { default as ZeroLimitAIIcon } from './zerolimitai'

components/icons/providers/zerolimitai.tsx

@@ -0,0 +1,25 @@
+import React from 'react'
+
+// Monochrome mark from https://www.zerolimitai.com (the "Z" route between two nodes)
+export default function ZeroLimitAIIcon({ className, ...props }: React.SVGProps<SVGSVGElement>) {
+  return (
+    <svg
+      xmlns="http://www.w3.org/2000/svg"
+      viewBox="0 0 24 24"
+      fill="none"
+      className={className}
+      aria-hidden="true"
+      {...props}
+    >
+      <path
+        d="M5 5.5H19L5 18.5H19"
+        stroke="currentColor"
+        strokeWidth="2.4"
+        strokeLinecap="round"
+        strokeLinejoin="round"
+      />
+      <circle cx="5" cy="5.5" r="3" fill="currentColor" />
+      <circle cx="19" cy="18.5" r="3" fill="currentColor" />
+    </svg>
+  )
+}

content/docs/configuration/librechat_yaml/ai_endpoints/meta.json

@@ -29,6 +29,7 @@
     "truefoundry",
     "vllm",
     "vultrcloudinference",
-    "xai"
+    "xai",
+    "zerolimitai"
   ]
 }

content/docs/configuration/librechat_yaml/ai_endpoints/zerolimitai.mdx

@@ -0,0 +1,38 @@
+---
+title: ZeroLimitAI
+icon: ZeroLimitAI
+description: Configure ZeroLimitAI as a custom endpoint in LibreChat.
+---
+
+ZeroLimitAI is an OpenAI-compatible endpoint that routes each request to the best free model available at that moment, with automatic failover when a model is rate-limited or down.
+
+## Get an API key
+
+Create a key on the [ZeroLimitAI developer page](https://www.zerolimitai.com/developer). Add it to your `.env` file:
+
+```bash filename=".env"
+ZEROLIMITAI_API_KEY=your-api-key
+```
+
+## Configuration
+
+Add the endpoint under `endpoints.custom` in your `librechat.yaml`:
+
+```yaml filename="librechat.yaml"
+    - name: "ZeroLimitAI"
+      apiKey: "${ZEROLIMITAI_API_KEY}"
+      baseURL: "https://www.zerolimitai.com/api/v1"
+      models:
+        default: ["auto"]
+        fetch: true
+      titleConvo: true
+      titleModel: "auto"
+      modelDisplayLabel: "ZeroLimitAI"
+```
+
+## Notes
+
+- `auto` lets the router choose the model for each request; the `model` field of the response names the model that answered. The live ranking is on the [leaderboard](https://www.zerolimitai.com/leaderboard).
+- With `fetch: true`, LibreChat lists the models currently in the ranking (`GET /v1/models`). The free catalogue changes over time, so `auto` is the stable choice.
+- Streaming and tool calls are supported; embeddings and image inputs are not.
+- A new key works for 7 days with 100 requests a day; after that the account needs a paid plan. See the [API reference](https://www.zerolimitai.com/docs/api) for limits and error codes.

lib/icons.tsx

@@ -52,6 +52,7 @@ import {
   VultrIcon,
   WolframIcon,
   XAIIcon,
+  ZeroLimitAIIcon,
 } from '../components/icons/providers'
 import {
   Rocket,
@@ -288,6 +289,7 @@ const icons: Record<string, ReactElement> = {
   Vultr: <VultrIcon />,
   Wolfram: <WolframIcon />,
   XAI: <XAIIcon />,
+  ZeroLimitAI: <ZeroLimitAIIcon />,
 }
 
 export function resolveIcon(icon: string | undefined): ReactElement | undefined {