fix: jan beta logo displayed in jan release (#4828)

chore: should have an option to not revalidate hardware information

.github/workflows/jan-electron-build.yml

@@ -40,6 +40,8 @@ jobs:
     with:
       ref: ${{ github.ref }}
       public_provider: github
+      beta: false
+      nightly: false
       new_version: ${{ needs.get-update-version.outputs.new_version }}
 
   build-windows-x64:
@@ -49,6 +51,8 @@ jobs:
     with:
       ref: ${{ github.ref }}
       public_provider: github
+      beta: false
+      nightly: false
       new_version: ${{ needs.get-update-version.outputs.new_version }}
 
   build-linux-x64:
@@ -58,6 +62,8 @@ jobs:
     with:
       ref: ${{ github.ref }}
       public_provider: github
+      beta: false
+      nightly: false
       new_version: ${{ needs.get-update-version.outputs.new_version }}
 
   update_release_draft:
@@ -82,4 +88,4 @@ jobs:
         #   config-name: my-config.yml
         #   disable-autolabeler: true
         env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

.github/workflows/nightly-integrate-cortex-cpp.yml

@@ -36,7 +36,7 @@ jobs:
       env:
         GITHUB_TOKEN: ${{ secrets.PAT_SERVICE_ACCOUNT }}
       run: |
-        curl -s https://api.github.com/repos/janhq/cortex/releases > /tmp/github_api_releases.json
+        curl -s https://api.github.com/repos/menloresearch/cortex/releases > /tmp/github_api_releases.json
         latest_prerelease_name=$(cat /tmp/github_api_releases.json | jq -r '.[] | select(.prerelease) | .name' | head -n 1)
 
         get_asset_count() {

.github/workflows/template-build-linux-x64.yml

@@ -55,55 +55,26 @@ jobs:
         if: inputs.beta == true && inputs.nightly != true
         shell: bash
         run: |
-          echo "Icons before replacement:"
+          rm -rf electron/icons/*
-          ls -la electron/icons/
+
-          
+          cp electron/icons_dev/jan-beta-512x512.png electron/icons/512x512.png
-          echo "Setting up beta icons"
+          cp electron/icons_dev/jan-beta.ico electron/icons/icon.ico
-          
+          cp electron/icons_dev/jan-beta.png electron/icons/icon.png
-          # Replace the key icon files with move operations (no need for rm first)
+          cp electron/icons_dev/jan-beta-tray@2x.png electron/icons/icon-tray@2x.png
-          mv electron/icons/jan-beta-512x512.png electron/icons/512x512.png
+          cp electron/icons_dev/jan-beta-tray.png electron/icons/icon-tray.png
-          mv electron/icons/jan-beta.ico electron/icons/icon.ico
-          mv electron/icons/jan-beta.png electron/icons/icon.png
-          mv electron/icons/jan-beta-tray@2x.png electron/icons/icon-tray@2x.png
-          mv electron/icons/jan-beta-tray.png electron/icons/icon-tray.png
-          
-          # Remove any remaining nightly icon files
-          rm -f electron/icons/jan-nightly-512x512.png
-          rm -f electron/icons/jan-nightly.ico
-          rm -f electron/icons/jan-nightly.png
-          rm -f electron/icons/jan-nightly-tray@2x.png
-          rm -f electron/icons/jan-nightly-tray.png
-          
-          # Verify the replacement
-          echo "Icons after replacement:"
-          ls -la electron/icons/
 
       - name: Replace Icons for Nightly Build
         if: inputs.nightly == true && inputs.beta != true
         shell: bash
         run: |
-          echo "Icons before replacement:"
+          rm -rf electron/icons/*
-          ls -la electron/icons/
+
-          
+          cp electron/icons_dev/jan-nightly-512x512.png electron/icons/512x512.png
-          echo "Setting up nightly icons"
+          cp electron/icons_dev/jan-nightly.ico electron/icons/icon.ico
-          
+          cp electron/icons_dev/jan-nightly.png electron/icons/icon.png
-          # Replace the key icon files with move operations (no need for rm first)
+          cp electron/icons_dev/jan-nightly-tray@2x.png electron/icons/icon-tray@2x.png
-          mv electron/icons/jan-nightly-512x512.png electron/icons/512x512.png
+          cp electron/icons_dev/jan-nightly-tray.png electron/icons/icon-tray.png
-          mv electron/icons/jan-nightly.ico electron/icons/icon.ico
+
-          mv electron/icons/jan-nightly.png electron/icons/icon.png
-          mv electron/icons/jan-nightly-tray@2x.png electron/icons/icon-tray@2x.png
-          mv electron/icons/jan-nightly-tray.png electron/icons/icon-tray.png
-          
-          # Remove any remaining beta icon files
-          rm -f electron/icons/jan-beta-512x512.png
-          rm -f electron/icons/jan-beta.ico
-          rm -f electron/icons/jan-beta.png
-          rm -f electron/icons/jan-beta-tray@2x.png
-          rm -f electron/icons/jan-beta-tray.png
-          
-          # Verify the replacement
-          echo "Icons after replacement:"
-          ls -la electron/icons/
 
       - name: Installing node
         uses: actions/setup-node@v1
@@ -213,4 +184,4 @@ jobs:
         uses: actions/upload-artifact@v4
         with:
           name: jan-linux-amd64-${{ inputs.new_version }}-AppImage
-          path: ./electron/dist/*.AppImage
+          path: ./electron/dist/*.AppImage

.github/workflows/template-build-macos.yml

@@ -65,31 +65,25 @@ jobs:
         if: inputs.beta == true && inputs.nightly != true
         shell: bash
         run: |
-          rm -f electron/icons/512x512.png
+          rm -rf electron/icons/*
-          rm -f electron/icons/icon.ico
+
-          rm -f electron/icons/icon.png
+          cp electron/icons_dev/jan-beta-512x512.png electron/icons/512x512.png
-          rm -f electron/icons/icon-tray@2x.png
+          cp electron/icons_dev/jan-beta.ico electron/icons/icon.ico
-          rm -f electron/icons/icon-tray.png
+          cp electron/icons_dev/jan-beta.png electron/icons/icon.png
-          mv electron/icons/jan-beta-512x512.png electron/icons/512x512.png
+          cp electron/icons_dev/jan-beta-tray@2x.png electron/icons/icon-tray@2x.png
-          mv electron/icons/jan-beta.ico electron/icons/icon.ico
+          cp electron/icons_dev/jan-beta-tray.png electron/icons/icon-tray.png
-          mv electron/icons/jan-beta.png electron/icons/icon.png
-          mv electron/icons/jan-beta-tray@2x.png electron/icons/icon-tray@2x.png
-          mv electron/icons/jan-beta-tray.png electron/icons/icon-tray.png
 
       - name: Replace Icons for Nightly Build
         if: inputs.nightly == true && inputs.beta != true
         shell: bash
         run: |
-          rm -f electron/icons/512x512.png
+          rm -rf electron/icons/*
-          rm -f electron/icons/icon.ico
+
-          rm -f electron/icons/icon.png
+          cp electron/icons_dev/jan-nightly-512x512.png electron/icons/512x512.png
-          rm -f electron/icons/icon-tray@2x.png
+          cp electron/icons_dev/jan-nightly.ico electron/icons/icon.ico
-          rm -f electron/icons/icon-tray.png
+          cp electron/icons_dev/jan-nightly.png electron/icons/icon.png
-          mv electron/icons/jan-nightly-512x512.png electron/icons/512x512.png
+          cp electron/icons_dev/jan-nightly-tray@2x.png electron/icons/icon-tray@2x.png
-          mv electron/icons/jan-nightly.ico electron/icons/icon.ico
+          cp electron/icons_dev/jan-nightly-tray.png electron/icons/icon-tray.png
-          mv electron/icons/jan-nightly.png electron/icons/icon.png
-          mv electron/icons/jan-nightly-tray@2x.png electron/icons/icon-tray@2x.png
-          mv electron/icons/jan-nightly-tray.png electron/icons/icon-tray.png
 
       - name: Installing node
         uses: actions/setup-node@v1
@@ -236,4 +230,4 @@ jobs:
         uses: actions/upload-artifact@v4
         with:
           name: jan-mac-universal-${{ inputs.new_version }}
-          path: ./electron/dist/*.dmg
+          path: ./electron/dist/*.dmg

.github/workflows/template-build-windows-x64.yml

@@ -64,31 +64,25 @@ jobs:
         if: inputs.beta == true && inputs.nightly != true
         shell: bash
         run: |
-          rm -f electron/icons/512x512.png
+          rm -rf electron/icons/*
-          rm -f electron/icons/icon.ico
+
-          rm -f electron/icons/icon.png
+          cp electron/icons_dev/jan-beta-512x512.png electron/icons/512x512.png
-          rm -f electron/icons/icon-tray@2x.png
+          cp electron/icons_dev/jan-beta.ico electron/icons/icon.ico
-          rm -f electron/icons/icon-tray.png
+          cp electron/icons_dev/jan-beta.png electron/icons/icon.png
-          mv electron/icons/jan-beta-512x512.png electron/icons/512x512.png
+          cp electron/icons_dev/jan-beta-tray@2x.png electron/icons/icon-tray@2x.png
-          mv electron/icons/jan-beta.ico electron/icons/icon.ico
+          cp electron/icons_dev/jan-beta-tray.png electron/icons/icon-tray.png
-          mv electron/icons/jan-beta.png electron/icons/icon.png
-          mv electron/icons/jan-beta-tray@2x.png electron/icons/icon-tray@2x.png
-          mv electron/icons/jan-beta-tray.png electron/icons/icon-tray.png
 
       - name: Replace Icons for Nightly Build
         if: inputs.nightly == true && inputs.beta != true
         shell: bash
         run: |
-          rm -f electron/icons/512x512.png
+          rm -rf electron/icons/*
-          rm -f electron/icons/icon.ico
+
-          rm -f electron/icons/icon.png
+          cp electron/icons_dev/jan-nightly-512x512.png electron/icons/512x512.png
-          rm -f electron/icons/icon-tray@2x.png
+          cp electron/icons_dev/jan-nightly.ico electron/icons/icon.ico
-          rm -f electron/icons/icon-tray.png
+          cp electron/icons_dev/jan-nightly.png electron/icons/icon.png
-          mv electron/icons/jan-nightly-512x512.png electron/icons/512x512.png
+          cp electron/icons_dev/jan-nightly-tray@2x.png electron/icons/icon-tray@2x.png
-          mv electron/icons/jan-nightly.ico electron/icons/icon.ico
+          cp electron/icons_dev/jan-nightly-tray.png electron/icons/icon-tray.png
-          mv electron/icons/jan-nightly.png electron/icons/icon.png
-          mv electron/icons/jan-nightly-tray@2x.png electron/icons/icon-tray@2x.png
-          mv electron/icons/jan-nightly-tray.png electron/icons/icon-tray.png
 
       - name: Installing node
         uses: actions/setup-node@v1
@@ -232,4 +226,4 @@ jobs:
         uses: actions/upload-artifact@v4
         with:
           name: jan-win-x64-${{ inputs.new_version }}
-          path: ./electron/dist/*.exe
+          path: ./electron/dist/*.exe

.github/workflows/template-get-update-version.yml

@@ -30,7 +30,7 @@ jobs:
           local max_retries=3
           local tag
           while [ $retries -lt $max_retries ]; do
-            tag=$(curl -s https://api.github.com/repos/janhq/jan/releases/latest | jq -r .tag_name)
+            tag=$(curl -s https://api.github.com/repos/menloresearch/jan/releases/latest | jq -r .tag_name)
             if [ -n "$tag" ] && [ "$tag" != "null" ]; then
               echo $tag
               return

core/src/types/api/index.ts

@@ -33,6 +33,8 @@ export enum NativeRoute {
   stopServer = 'stopServer',
 
   appUpdateDownload = 'appUpdateDownload',
+
+  appToken = 'appToken',
 }
 
 /**

electron/handlers/native.ts

@@ -317,4 +317,11 @@ export function handleAppIPCs() {
     const { stopServer } = require('@janhq/server')
     return stopServer()
   })
+
+  /**
+   * Handles the "appToken" IPC message to generate a random app ID.
+   */
+  ipcMain.handle(NativeRoute.appToken, async (_event): Promise<string> => {
+    return process.env.appToken ?? 'cortex.cpp'
+  })
 }

electron/main.ts

@@ -29,6 +29,7 @@ import { trayManager } from './managers/tray'
 import { logSystemInfo } from './utils/system'
 import { registerGlobalShortcuts } from './utils/shortcut'
 import { registerLogger } from './utils/logger'
+import { randomBytes } from 'crypto'
 
 const preloadPath = join(__dirname, 'preload.js')
 const preloadQuickAskPath = join(__dirname, 'preload.quickask.js')
@@ -56,6 +57,10 @@ const createMainWindow = () => {
   windowManager.createMainWindow(preloadPath, startUrl)
 }
 
+// Generate a random token for the app
+// This token is used for authentication when making request to cortex.cpp server
+process.env.appToken = randomBytes(16).toString('hex')
+
 app
   .whenReady()
   .then(() => {

extensions/assistant-extension/src/node/retrieval.ts

@@ -23,11 +23,16 @@ export class Retrieval {
 
   constructor(chunkSize: number = 4000, chunkOverlap: number = 200) {
     this.updateTextSplitter(chunkSize, chunkOverlap)
+    this.initialize()
+  }
+
+  private async initialize() {
+    const apiKey = await window.core?.api.appToken() ?? 'cortex.cpp'
 
     // declare time-weighted retriever and storage
     this.timeWeightedVectorStore = new MemoryVectorStore(
       new OpenAIEmbeddings(
-        { openAIApiKey: 'cortex-embedding' },
+        { openAIApiKey: apiKey },
         { basePath: `${CORTEX_API_URL}/v1` }
       )
     )
@@ -47,9 +52,10 @@ export class Retrieval {
     })
   }
 
-  public updateEmbeddingEngine(model: string, engine: string): void {
+  public async updateEmbeddingEngine(model: string, engine: string) {
+    const apiKey = await window.core?.api.appToken() ?? 'cortex.cpp'
     this.embeddingModel = new OpenAIEmbeddings(
-      { openAIApiKey: 'cortex-embedding', model },
+      { openAIApiKey: apiKey, model },
       // TODO: Raw settings
       { basePath: `${CORTEX_API_URL}/v1` }
     )

extensions/conversational-extension/src/index.ts

@@ -4,7 +4,7 @@ import {
   ThreadAssistantInfo,
   ThreadMessage,
 } from '@janhq/core'
-import ky from 'ky'
+import ky, { KyInstance } from 'ky'
 import PQueue from 'p-queue'
 
 type ThreadList = {
@@ -22,6 +22,22 @@ type MessageList = {
 export default class CortexConversationalExtension extends ConversationalExtension {
   queue = new PQueue({ concurrency: 1 })
 
+  api?: KyInstance
+  /**
+   * Get the API instance
+   * @returns
+   */
+  async apiInstance(): Promise<KyInstance> {
+    if(this.api) return this.api
+    const apiKey = (await window.core?.api.appToken()) ?? 'cortex.cpp'
+    this.api = ky.extend({
+      prefixUrl: API_URL,
+      headers: {
+        Authorization: `Bearer ${apiKey}`,
+      },
+    })
+    return this.api
+  }
   /**
    * Called when the extension is loaded.
    */
@@ -39,10 +55,12 @@ export default class CortexConversationalExtension extends ConversationalExtensi
    */
   async listThreads(): Promise<Thread[]> {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .get(`${API_URL}/v1/threads?limit=-1`)
+        api
-        .json<ThreadList>()
+          .get('v1/threads?limit=-1')
-        .then((e) => e.data)
+          .json<ThreadList>()
+          .then((e) => e.data)
+      )
     ) as Promise<Thread[]>
   }
 
@@ -52,7 +70,9 @@ export default class CortexConversationalExtension extends ConversationalExtensi
    */
   async createThread(thread: Thread): Promise<Thread> {
     return this.queue.add(() =>
-      ky.post(`${API_URL}/v1/threads`, { json: thread }).json<Thread>()
+      this.apiInstance().then((api) =>
+        api.post('v1/threads', { json: thread }).json<Thread>()
+      )
     ) as Promise<Thread>
   }
 
@@ -63,7 +83,9 @@ export default class CortexConversationalExtension extends ConversationalExtensi
   async modifyThread(thread: Thread): Promise<void> {
     return this.queue
       .add(() =>
-        ky.patch(`${API_URL}/v1/threads/${thread.id}`, { json: thread })
+        this.apiInstance().then((api) =>
+          api.patch(`v1/threads/${thread.id}`, { json: thread })
+        )
       )
       .then()
   }
@@ -74,7 +96,9 @@ export default class CortexConversationalExtension extends ConversationalExtensi
    */
   async deleteThread(threadId: string): Promise<void> {
     return this.queue
-      .add(() => ky.delete(`${API_URL}/v1/threads/${threadId}`))
+      .add(() =>
+        this.apiInstance().then((api) => api.delete(`v1/threads/${threadId}`))
+      )
       .then()
   }
 
@@ -85,11 +109,13 @@ export default class CortexConversationalExtension extends ConversationalExtensi
    */
   async createMessage(message: ThreadMessage): Promise<ThreadMessage> {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .post(`${API_URL}/v1/threads/${message.thread_id}/messages`, {
+        api
-          json: message,
+          .post(`v1/threads/${message.thread_id}/messages`, {
-        })
+            json: message,
-        .json<ThreadMessage>()
+          })
+          .json<ThreadMessage>()
+      )
     ) as Promise<ThreadMessage>
   }
 
@@ -100,14 +126,13 @@ export default class CortexConversationalExtension extends ConversationalExtensi
    */
   async modifyMessage(message: ThreadMessage): Promise<ThreadMessage> {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .patch(
+        api
-          `${API_URL}/v1/threads/${message.thread_id}/messages/${message.id}`,
+          .patch(`v1/threads/${message.thread_id}/messages/${message.id}`, {
-          {
             json: message,
-          }
+          })
-        )
+          .json<ThreadMessage>()
-        .json<ThreadMessage>()
+      )
     ) as Promise<ThreadMessage>
   }
 
@@ -120,7 +145,9 @@ export default class CortexConversationalExtension extends ConversationalExtensi
   async deleteMessage(threadId: string, messageId: string): Promise<void> {
     return this.queue
       .add(() =>
-        ky.delete(`${API_URL}/v1/threads/${threadId}/messages/${messageId}`)
+        this.apiInstance().then((api) =>
+          api.delete(`v1/threads/${threadId}/messages/${messageId}`)
+        )
       )
       .then()
   }
@@ -132,10 +159,12 @@ export default class CortexConversationalExtension extends ConversationalExtensi
    */
   async listMessages(threadId: string): Promise<ThreadMessage[]> {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .get(`${API_URL}/v1/threads/${threadId}/messages?order=asc&limit=-1`)
+        api
-        .json<MessageList>()
+          .get(`v1/threads/${threadId}/messages?order=asc&limit=-1`)
-        .then((e) => e.data)
+          .json<MessageList>()
+          .then((e) => e.data)
+      )
     ) as Promise<ThreadMessage[]>
   }
 
@@ -147,9 +176,11 @@ export default class CortexConversationalExtension extends ConversationalExtensi
    */
   async getThreadAssistant(threadId: string): Promise<ThreadAssistantInfo> {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .get(`${API_URL}/v1/assistants/${threadId}?limit=-1`)
+        api
-        .json<ThreadAssistantInfo>()
+          .get(`v1/assistants/${threadId}?limit=-1`)
+          .json<ThreadAssistantInfo>()
+      )
     ) as Promise<ThreadAssistantInfo>
   }
   /**
@@ -163,9 +194,11 @@ export default class CortexConversationalExtension extends ConversationalExtensi
     assistant: ThreadAssistantInfo
   ): Promise<ThreadAssistantInfo> {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .post(`${API_URL}/v1/assistants/${threadId}`, { json: assistant })
+        api
-        .json<ThreadAssistantInfo>()
+          .post(`v1/assistants/${threadId}`, { json: assistant })
+          .json<ThreadAssistantInfo>()
+      )
     ) as Promise<ThreadAssistantInfo>
   }
 
@@ -180,9 +213,11 @@ export default class CortexConversationalExtension extends ConversationalExtensi
     assistant: ThreadAssistantInfo
   ): Promise<ThreadAssistantInfo> {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .patch(`${API_URL}/v1/assistants/${threadId}`, { json: assistant })
+        api
-        .json<ThreadAssistantInfo>()
+          .patch(`v1/assistants/${threadId}`, { json: assistant })
+          .json<ThreadAssistantInfo>()
+      )
     ) as Promise<ThreadAssistantInfo>
   }
 
@@ -191,10 +226,12 @@ export default class CortexConversationalExtension extends ConversationalExtensi
    * @returns
    */
   async healthz(): Promise<void> {
-    return ky
+    return this.apiInstance()
-      .get(`${API_URL}/healthz`, {
+      .then((api) =>
-        retry: { limit: 20, delay: () => 500, methods: ['get'] },
+        api.get('healthz', {
-      })
+          retry: { limit: 20, delay: () => 500, methods: ['get'] },
+        })
+      )
       .then(() => {})
   }
 }

extensions/engine-management-extension/src/index.ts

@@ -15,7 +15,7 @@ import {
   ModelEvent,
   EngineEvent,
 } from '@janhq/core'
-import ky, { HTTPError } from 'ky'
+import ky, { HTTPError, KyInstance } from 'ky'
 import PQueue from 'p-queue'
 import { EngineError } from './error'
 import { getJanDataFolderPath } from '@janhq/core'
@@ -31,6 +31,22 @@ interface ModelList {
 export default class JanEngineManagementExtension extends EngineManagementExtension {
   queue = new PQueue({ concurrency: 1 })
 
+  api?: KyInstance
+  /**
+   * Get the API instance
+   * @returns
+   */
+  async apiInstance(): Promise<KyInstance> {
+    if(this.api) return this.api
+    const apiKey = (await window.core?.api.appToken()) ?? 'cortex.cpp'
+    this.api = ky.extend({
+      prefixUrl: API_URL,
+      headers: {
+        Authorization: `Bearer ${apiKey}`,
+      },
+    })
+    return this.api
+  }
   /**
    * Called when the extension is loaded.
    */
@@ -59,10 +75,12 @@ export default class JanEngineManagementExtension extends EngineManagementExtens
    */
   async getEngines(): Promise<Engines> {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .get(`${API_URL}/v1/engines`)
+        api
-        .json<Engines>()
+          .get('v1/engines')
-        .then((e) => e)
+          .json<Engines>()
+          .then((e) => e)
+      )
     ) as Promise<Engines>
   }
 
@@ -70,12 +88,15 @@ export default class JanEngineManagementExtension extends EngineManagementExtens
    * @returns A Promise that resolves to an object of list engines.
    */
   async getRemoteModels(name: string): Promise<any> {
-    return ky
+    return this.apiInstance().then(
-      .get(`${API_URL}/v1/models/remote/${name}`)
+      (api) =>
-      .json<ModelList>()
+        api
-      .catch(() => ({
+          .get(`v1/models/remote/${name}`)
-        data: [],
+          .json<ModelList>()
-      })) as Promise<ModelList>
+          .catch(() => ({
+            data: [],
+          })) as Promise<ModelList>
+    )
   }
 
   /**
@@ -84,10 +105,12 @@ export default class JanEngineManagementExtension extends EngineManagementExtens
    */
   async getInstalledEngines(name: InferenceEngine): Promise<EngineVariant[]> {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .get(`${API_URL}/v1/engines/${name}`)
+        api
-        .json<EngineVariant[]>()
+          .get(`v1/engines/${name}`)
-        .then((e) => e)
+          .json<EngineVariant[]>()
+          .then((e) => e)
+      )
     ) as Promise<EngineVariant[]>
   }
 
@@ -103,12 +126,14 @@ export default class JanEngineManagementExtension extends EngineManagementExtens
     platform?: string
   ) {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .get(`${API_URL}/v1/engines/${name}/releases/${version}`)
+        api
-        .json<EngineReleased[]>()
+          .get(`v1/engines/${name}/releases/${version}`)
-        .then((e) =>
+          .json<EngineReleased[]>()
-          platform ? e.filter((r) => r.name.includes(platform)) : e
+          .then((e) =>
-        )
+            platform ? e.filter((r) => r.name.includes(platform)) : e
+          )
+      )
     ) as Promise<EngineReleased[]>
   }
 
@@ -119,12 +144,14 @@ export default class JanEngineManagementExtension extends EngineManagementExtens
    */
   async getLatestReleasedEngine(name: InferenceEngine, platform?: string) {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .get(`${API_URL}/v1/engines/${name}/releases/latest`)
+        api
-        .json<EngineReleased[]>()
+          .get(`v1/engines/${name}/releases/latest`)
-        .then((e) =>
+          .json<EngineReleased[]>()
-          platform ? e.filter((r) => r.name.includes(platform)) : e
+          .then((e) =>
-        )
+            platform ? e.filter((r) => r.name.includes(platform)) : e
+          )
+      )
     ) as Promise<EngineReleased[]>
   }
 
@@ -134,9 +161,11 @@ export default class JanEngineManagementExtension extends EngineManagementExtens
    */
   async installEngine(name: string, engineConfig: EngineConfig) {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .post(`${API_URL}/v1/engines/${name}/install`, { json: engineConfig })
+        api
-        .then((e) => e)
+          .post(`v1/engines/${name}/install`, { json: engineConfig })
+          .then((e) => e)
+      )
     ) as Promise<{ messages: string }>
   }
 
@@ -167,15 +196,17 @@ export default class JanEngineManagementExtension extends EngineManagementExtens
       engineConfig.metadata.header_template = DEFAULT_REQUEST_HEADERS_TRANSFORM
 
     return this.queue.add(() =>
-      ky.post(`${API_URL}/v1/engines`, { json: engineConfig }).then((e) => {
+      this.apiInstance().then((api) =>
-        if (persistModels && engineConfig.metadata?.get_models_url) {
+        api.post('v1/engines', { json: engineConfig }).then((e) => {
-          // Pull /models from remote models endpoint
+          if (persistModels && engineConfig.metadata?.get_models_url) {
-          return this.populateRemoteModels(engineConfig)
+            // Pull /models from remote models endpoint
-            .then(() => e)
+            return this.populateRemoteModels(engineConfig)
-            .catch(() => e)
+              .then(() => e)
-        }
+              .catch(() => e)
-        return e
+          }
-      })
+          return e
+        })
+      )
     ) as Promise<{ messages: string }>
   }
 
@@ -185,9 +216,11 @@ export default class JanEngineManagementExtension extends EngineManagementExtens
    */
   async uninstallEngine(name: InferenceEngine, engineConfig: EngineConfig) {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .delete(`${API_URL}/v1/engines/${name}/install`, { json: engineConfig })
+        api
-        .then((e) => e)
+          .delete(`v1/engines/${name}/install`, { json: engineConfig })
+          .then((e) => e)
+      )
     ) as Promise<{ messages: string }>
   }
 
@@ -196,25 +229,27 @@ export default class JanEngineManagementExtension extends EngineManagementExtens
    * @param model - Remote model object.
    */
   async addRemoteModel(model: Model) {
-    return this.queue
+    return this.queue.add(() =>
-      .add(() =>
+      this.apiInstance()
-        ky
+        .then((api) =>
-          .post(`${API_URL}/v1/models/add`, {
+          api
-            json: {
+            .post('v1/models/add', {
-              inference_params: {
+              json: {
-                max_tokens: 4096,
+                inference_params: {
-                temperature: 0.7,
+                  max_tokens: 4096,
-                top_p: 0.95,
+                  temperature: 0.7,
-                stream: true,
+                  top_p: 0.95,
-                frequency_penalty: 0,
+                  stream: true,
-                presence_penalty: 0,
+                  frequency_penalty: 0,
+                  presence_penalty: 0,
+                },
+                ...model,
               },
-              ...model,
+            })
-            },
+            .then((e) => e)
-          })
+        )
-          .then((e) => e)
+        .then(() => {})
-      )
+    )
-      .then(() => {})
   }
 
   /**
@@ -223,10 +258,12 @@ export default class JanEngineManagementExtension extends EngineManagementExtens
    */
   async getDefaultEngineVariant(name: InferenceEngine) {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .get(`${API_URL}/v1/engines/${name}/default`)
+        api
-        .json<{ messages: string }>()
+          .get(`v1/engines/${name}/default`)
-        .then((e) => e)
+          .json<{ messages: string }>()
+          .then((e) => e)
+      )
     ) as Promise<DefaultEngineVariant>
   }
 
@@ -240,9 +277,11 @@ export default class JanEngineManagementExtension extends EngineManagementExtens
     engineConfig: EngineConfig
   ) {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .post(`${API_URL}/v1/engines/${name}/default`, { json: engineConfig })
+        api
-        .then((e) => e)
+          .post(`v1/engines/${name}/default`, { json: engineConfig })
+          .then((e) => e)
+      )
     ) as Promise<{ messages: string }>
   }
 
@@ -251,9 +290,11 @@ export default class JanEngineManagementExtension extends EngineManagementExtens
    */
   async updateEngine(name: InferenceEngine, engineConfig?: EngineConfig) {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .post(`${API_URL}/v1/engines/${name}/update`, { json: engineConfig })
+        api
-        .then((e) => e)
+          .post(`v1/engines/${name}/update`, { json: engineConfig })
+          .then((e) => e)
+      )
     ) as Promise<{ messages: string }>
   }
 
@@ -262,10 +303,12 @@ export default class JanEngineManagementExtension extends EngineManagementExtens
    * @returns
    */
   async healthz(): Promise<void> {
-    return ky
+    return this.apiInstance()
-      .get(`${API_URL}/healthz`, {
+      .then((api) =>
-        retry: { limit: 20, delay: () => 500, methods: ['get'] },
+        api.get('healthz', {
-      })
+          retry: { limit: 20, delay: () => 500, methods: ['get'] },
+        })
+      )
       .then(() => {
         this.queue.concurrency = Infinity
       })
@@ -390,7 +433,6 @@ export default class JanEngineManagementExtension extends EngineManagementExtens
     const version = await this.getSetting<string>('version', '0.0.0')
     const engines = await this.getEngines()
     if (version < VERSION) {
-
       console.log('Migrating engine settings...')
       // Migrate engine settings
       await Promise.all(
@@ -398,7 +440,7 @@ export default class JanEngineManagementExtension extends EngineManagementExtens
           const { id, ...data } = engine
 
           data.api_key = engines[id]?.api_key
-          return this.updateEngine(id,{
+          return this.updateEngine(id, {
             ...data,
           }).catch(console.error)
         })

extensions/hardware-management-extension/package.json

@@ -29,12 +29,10 @@
   },
   "dependencies": {
     "@janhq/core": "../../core/package.tgz",
-    "cpu-instructions": "^0.0.13",
     "ky": "^1.7.2",
     "p-queue": "^8.0.1"
   },
   "bundledDependencies": [
-    "cpu-instructions",
     "@janhq/core"
   ],
   "hardwares": {

extensions/hardware-management-extension/src/index.ts

@@ -1,5 +1,5 @@
 import { HardwareManagementExtension, HardwareInformation } from '@janhq/core'
-import ky from 'ky'
+import ky, { KyInstance } from 'ky'
 import PQueue from 'p-queue'
 
 /**
@@ -17,6 +17,23 @@ export default class JSONHardwareManagementExtension extends HardwareManagementE
     this.queue.add(() => this.healthz())
   }
 
+  api?: KyInstance
+  /**
+   * Get the API instance
+   * @returns
+   */
+  async apiInstance(): Promise<KyInstance> {
+    if(this.api) return this.api
+    const apiKey = (await window.core?.api.appToken()) ?? 'cortex.cpp'
+    this.api = ky.extend({
+      prefixUrl: API_URL,
+      headers: {
+        Authorization: `Bearer ${apiKey}`,
+      },
+    })
+    return this.api
+  }
+
   /**
    * Called when the extension is unloaded.
    */
@@ -27,11 +44,13 @@ export default class JSONHardwareManagementExtension extends HardwareManagementE
    * @returns
    */
   async healthz(): Promise<void> {
-    return ky
+    return this.apiInstance().then((api) =>
-      .get(`${API_URL}/healthz`, {
+      api
-        retry: { limit: 20, delay: () => 500, methods: ['get'] },
+        .get('healthz', {
-      })
+          retry: { limit: 20, delay: () => 500, methods: ['get'] },
-      .then(() => {})
+        })
+        .then(() => {})
+    )
   }
 
   /**
@@ -39,10 +58,12 @@ export default class JSONHardwareManagementExtension extends HardwareManagementE
    */
   async getHardware(): Promise<HardwareInformation> {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .get(`${API_URL}/v1/hardware`)
+        api
-        .json<HardwareInformation>()
+          .get('v1/hardware')
-        .then((e) => e)
+          .json<HardwareInformation>()
+          .then((e) => e)
+      )
     ) as Promise<HardwareInformation>
   }
 
@@ -54,7 +75,9 @@ export default class JSONHardwareManagementExtension extends HardwareManagementE
     activated_gpus: number[]
   }> {
     return this.queue.add(() =>
-      ky.post(`${API_URL}/v1/hardware/activate`, { json: data }).then((e) => e)
+      this.apiInstance().then((api) =>
+        api.post('v1/hardware/activate', { json: data }).then((e) => e)
+      )
     ) as Promise<{
       message: string
       activated_gpus: number[]

extensions/inference-cortex-extension/bin/version.txt

@@ -1 +1 @@
-1.0.11-rc9
+1.0.12

extensions/inference-cortex-extension/download.bat

@@ -5,11 +5,11 @@ set /p CORTEX_VERSION=<./bin/version.txt
 set ENGINE_VERSION=0.1.55
 
 @REM Download cortex.llamacpp binaries
-set DOWNLOAD_URL=https://github.com/janhq/cortex.llamacpp/releases/download/v%ENGINE_VERSION%/cortex.llamacpp-%ENGINE_VERSION%-windows-amd64
+set DOWNLOAD_URL=https://github.com/menloresearch/cortex.llamacpp/releases/download/v%ENGINE_VERSION%/cortex.llamacpp-%ENGINE_VERSION%-windows-amd64
-set CUDA_DOWNLOAD_URL=https://github.com/janhq/cortex.llamacpp/releases/download/v%ENGINE_VERSION%
+set CUDA_DOWNLOAD_URL=https://github.com/menloresearch/cortex.llamacpp/releases/download/v%ENGINE_VERSION%
 set SUBFOLDERS=windows-amd64-noavx-cuda-12-0 windows-amd64-noavx-cuda-11-7 windows-amd64-avx2-cuda-12-0 windows-amd64-avx2-cuda-11-7 windows-amd64-noavx windows-amd64-avx windows-amd64-avx2 windows-amd64-avx512 windows-amd64-vulkan
 
-call .\node_modules\.bin\download -e --strip 1 -o %BIN_PATH% https://github.com/janhq/cortex.cpp/releases/download/v%CORTEX_VERSION%/cortex-%CORTEX_VERSION%-windows-amd64.tar.gz
+call .\node_modules\.bin\download -e --strip 1 -o %BIN_PATH% https://github.com/menloresearch/cortex.cpp/releases/download/v%CORTEX_VERSION%/cortex-%CORTEX_VERSION%-windows-amd64.tar.gz
 call .\node_modules\.bin\download %DOWNLOAD_URL%-avx2-cuda-12-0.tar.gz -e --strip 1 -o %SHARED_PATH%/engines/cortex.llamacpp/windows-amd64-avx2-cuda-12-0/v%ENGINE_VERSION%
 call .\node_modules\.bin\download %DOWNLOAD_URL%-avx2-cuda-11-7.tar.gz -e --strip 1 -o %SHARED_PATH%/engines/cortex.llamacpp/windows-amd64-avx2-cuda-11-7/v%ENGINE_VERSION%
 call .\node_modules\.bin\download %DOWNLOAD_URL%-noavx-cuda-12-0.tar.gz -e --strip 1 -o %SHARED_PATH%/engines/cortex.llamacpp/windows-amd64-noavx-cuda-12-0/v%ENGINE_VERSION%

extensions/inference-cortex-extension/download.sh

@@ -3,9 +3,9 @@
 # Read CORTEX_VERSION
 CORTEX_VERSION=$(cat ./bin/version.txt)
 ENGINE_VERSION=0.1.55
-CORTEX_RELEASE_URL="https://github.com/janhq/cortex.cpp/releases/download"
+CORTEX_RELEASE_URL="https://github.com/menloresearch/cortex.cpp/releases/download"
-ENGINE_DOWNLOAD_URL="https://github.com/janhq/cortex.llamacpp/releases/download/v${ENGINE_VERSION}/cortex.llamacpp-${ENGINE_VERSION}"
+ENGINE_DOWNLOAD_URL="https://github.com/menloresearch/cortex.llamacpp/releases/download/v${ENGINE_VERSION}/cortex.llamacpp-${ENGINE_VERSION}"
-CUDA_DOWNLOAD_URL="https://github.com/janhq/cortex.llamacpp/releases/download/v${ENGINE_VERSION}"
+CUDA_DOWNLOAD_URL="https://github.com/menloresearch/cortex.llamacpp/releases/download/v${ENGINE_VERSION}"
 BIN_PATH=./bin
 SHARED_PATH="../../electron/shared"
 # Detect platform

extensions/inference-cortex-extension/rolldown.config.mjs

@@ -24,7 +24,7 @@ export default defineConfig([
   },
   {
     input: 'src/node/index.ts',
-    external: ['@janhq/core/node', 'cpu-instructions'],
+    external: ['@janhq/core/node'],
     output: {
       format: 'cjs',
       file: 'dist/node/index.cjs.js',

extensions/inference-cortex-extension/src/index.ts

@@ -17,7 +17,7 @@ import {
   ModelEvent,
 } from '@janhq/core'
 import PQueue from 'p-queue'
-import ky from 'ky'
+import ky, { KyInstance } from 'ky'
 
 /**
  * Event subscription types of Downloader
@@ -75,8 +75,35 @@ export default class JanInferenceCortexExtension extends LocalOAIEngine {
 
   abortControllers = new Map<string, AbortController>()
 
+  api?: KyInstance
   /**
-   * Subscribes to events emitted by the @janhq/core package.
+   * Get the API instance
+   * @returns
+   */
+  async apiInstance(): Promise<KyInstance> {
+    if(this.api) return this.api
+    const apiKey = (await window.core?.api.appToken()) ?? 'cortex.cpp'
+    this.api = ky.extend({
+      prefixUrl: CORTEX_API_URL,
+      headers: {
+        Authorization: `Bearer ${apiKey}`,
+      },
+    })
+    return this.api
+  }
+
+  /**
+   * Authorization headers for the API requests.
+   * @returns
+   */
+  headers(): Promise<HeadersInit> {
+    return window.core?.api.appToken().then((token: string) => ({
+      Authorization: `Bearer ${token}`,
+    }))
+  }
+
+  /**
+   * Called when the extension is loaded.
    */
   async onLoad() {
     super.onLoad()
@@ -153,45 +180,49 @@ export default class JanInferenceCortexExtension extends LocalOAIEngine {
     this.abortControllers.set(model.id, controller)
 
     return await this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .post(`${CORTEX_API_URL}/v1/models/start`, {
+        api
-          json: {
+          .post('v1/models/start', {
-            ...extractModelLoadParams(model.settings),
+            json: {
-            model: model.id,
+              ...extractModelLoadParams(model.settings),
-            engine:
+              model: model.id,
-              model.engine === InferenceEngine.nitro // Legacy model cache
+              engine:
-                ? InferenceEngine.cortex_llamacpp
+                model.engine === InferenceEngine.nitro // Legacy model cache
-                : model.engine,
+                  ? InferenceEngine.cortex_llamacpp
-            cont_batching: this.cont_batching,
+                  : model.engine,
-            n_parallel: this.n_parallel,
+              cont_batching: this.cont_batching,
-            caching_enabled: this.caching_enabled,
+              n_parallel: this.n_parallel,
-            flash_attn: this.flash_attn,
+              caching_enabled: this.caching_enabled,
-            cache_type: this.cache_type,
+              flash_attn: this.flash_attn,
-            use_mmap: this.use_mmap,
+              cache_type: this.cache_type,
-            ...(this.cpu_threads ? { cpu_threads: this.cpu_threads } : {}),
+              use_mmap: this.use_mmap,
-          },
+              ...(this.cpu_threads ? { cpu_threads: this.cpu_threads } : {}),
-          timeout: false,
+            },
-          signal,
+            timeout: false,
-        })
+            signal,
-        .json()
+          })
-        .catch(async (e) => {
+          .json()
-          throw (await e.response?.json()) ?? e
+          .catch(async (e) => {
-        })
+            throw (await e.response?.json()) ?? e
-        .finally(() => this.abortControllers.delete(model.id))
+          })
-        .then()
+          .finally(() => this.abortControllers.delete(model.id))
+          .then()
+      )
     )
   }
 
   override async unloadModel(model: Model): Promise<void> {
-    return ky
+    return this.apiInstance().then((api) =>
-      .post(`${CORTEX_API_URL}/v1/models/stop`, {
+      api
-        json: { model: model.id },
+        .post('v1/models/stop', {
-      })
+          json: { model: model.id },
-      .json()
+        })
-      .finally(() => {
+        .json()
-        this.abortControllers.get(model.id)?.abort()
+        .finally(() => {
-      })
+          this.abortControllers.get(model.id)?.abort()
-      .then()
+        })
+        .then()
+    )
   }
 
   /**
@@ -199,15 +230,17 @@ export default class JanInferenceCortexExtension extends LocalOAIEngine {
    * @returns
    */
   private async healthz(): Promise<void> {
-    return ky
+    return this.apiInstance().then((api) =>
-      .get(`${CORTEX_API_URL}/healthz`, {
+      api
-        retry: {
+        .get('healthz', {
-          limit: 20,
+          retry: {
-          delay: () => 500,
+            limit: 20,
-          methods: ['get'],
+            delay: () => 500,
-        },
+            methods: ['get'],
-      })
+          },
-      .then(() => {})
+        })
+        .then(() => {})
+    )
   }
 
   /**
@@ -215,13 +248,15 @@ export default class JanInferenceCortexExtension extends LocalOAIEngine {
    * @returns
    */
   private async clean(): Promise<any> {
-    return ky
+    return this.apiInstance()
-      .delete(`${CORTEX_API_URL}/processmanager/destroy`, {
+      .then((api) =>
-        timeout: 2000, // maximum 2 seconds
+        api.delete('processmanager/destroy', {
-        retry: {
+          timeout: 2000, // maximum 2 seconds
-          limit: 0,
+          retry: {
-        },
+            limit: 0,
-      })
+          },
+        })
+      )
       .catch(() => {
         // Do nothing
       })

extensions/inference-cortex-extension/src/node/index.ts

@@ -44,8 +44,9 @@ function run(): Promise<any> {
         `${path.join(dataFolderPath, '.janrc')}`,
         '--data_folder_path',
         dataFolderPath,
-        '--loglevel',
+        'config',
-        'INFO',
+        '--api_keys',
+        process.env.appToken ?? 'cortex.cpp',
       ],
       {
         env: {

extensions/model-extension/rolldown.config.mjs

@@ -11,7 +11,7 @@ export default defineConfig({
   platform: 'browser',
   define: {
     SETTINGS: JSON.stringify(settingJson),
-    API_URL: JSON.stringify(`http://127.0.0.1:${process.env.CORTEX_API_PORT ?? "39291"}`),
+    CORTEX_API_URL: JSON.stringify(`http://127.0.0.1:${process.env.CORTEX_API_PORT ?? "39291"}`),
     DEFAULT_MODEL_SOURCES: JSON.stringify(modelSources),
   },
 })

extensions/model-extension/src/@types/global.d.ts

@@ -1,5 +1,5 @@
 declare const NODE: string
-declare const API_URL: string
+declare const CORTEX_API_URL: string
 declare const SETTINGS: SettingComponentProps[]
 declare const DEFAULT_MODEL_SOURCES: any
 

extensions/model-extension/src/index.ts

@@ -13,7 +13,7 @@ import {
 import { scanModelsFolder } from './legacy/model-json'
 import { deleteModelFiles } from './legacy/delete'
 import PQueue from 'p-queue'
-import ky from 'ky'
+import ky, { KyInstance } from 'ky'
 
 /**
  * cortex.cpp setting keys
@@ -32,9 +32,25 @@ type Data<T> = {
  */
 export default class JanModelExtension extends ModelExtension {
   queue = new PQueue({ concurrency: 1 })
+
+  api?: KyInstance
+  /**
+   * Get the API instance
+   * @returns
+   */
+  async apiInstance(): Promise<KyInstance> {
+    if(this.api) return this.api
+    const apiKey = (await window.core?.api.appToken()) ?? 'cortex.cpp'
+    this.api = ky.extend({
+      prefixUrl: CORTEX_API_URL,
+      headers: {
+        Authorization: `Bearer ${apiKey}`,
+      },
+    })
+    return this.api
+  }
   /**
    * Called when the extension is loaded.
-   * @override
    */
   async onLoad() {
     this.queue.add(() => this.healthz())
@@ -82,13 +98,15 @@ export default class JanModelExtension extends ModelExtension {
      * Sending POST to /models/pull/{id} endpoint to pull the model
      */
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .post(`${API_URL}/v1/models/pull`, { json: { model, id, name } })
+        api
-        .json()
+          .post('v1/models/pull', { json: { model, id, name }, timeout: false })
-        .catch(async (e) => {
+          .json()
-          throw (await e.response?.json()) ?? e
+          .catch(async (e) => {
-        })
+            throw (await e.response?.json()) ?? e
-        .then()
+          })
+          .then()
+      )
     )
   }
 
@@ -103,10 +121,12 @@ export default class JanModelExtension extends ModelExtension {
      * Sending DELETE to /models/pull/{id} endpoint to cancel a model pull
      */
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .delete(`${API_URL}/v1/models/pull`, { json: { taskId: model } })
+        api
-        .json()
+          .delete('v1/models/pull', { json: { taskId: model } })
-        .then()
+          .json()
+          .then()
+      )
     )
   }
 
@@ -117,7 +137,11 @@ export default class JanModelExtension extends ModelExtension {
    */
   async deleteModel(model: string): Promise<void> {
     return this.queue
-      .add(() => ky.delete(`${API_URL}/v1/models/${model}`).json().then())
+      .add(() =>
+        this.apiInstance().then((api) =>
+          api.delete(`v1/models/${model}`).json().then()
+        )
+      )
       .catch((e) => console.debug(e))
       .finally(async () => {
         // Delete legacy model files
@@ -219,10 +243,15 @@ export default class JanModelExtension extends ModelExtension {
   async updateModel(model: Partial<Model>): Promise<Model> {
     return this.queue
       .add(() =>
-        ky
+        this.apiInstance().then((api) =>
-          .patch(`${API_URL}/v1/models/${model.id}`, { json: { ...model } })
+          api
-          .json()
+            .patch(`v1/models/${model.id}`, {
-          .then()
+              json: { ...model },
+              timeout: false,
+            })
+            .json()
+            .then()
+        )
       )
       .then(() => this.getModel(model.id))
   }
@@ -233,10 +262,12 @@ export default class JanModelExtension extends ModelExtension {
    */
   async getModel(model: string): Promise<Model> {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .get(`${API_URL}/v1/models/${model}`)
+        api
-        .json()
+          .get(`v1/models/${model}`)
-        .then((e) => this.transformModel(e))
+          .json()
+          .then((e) => this.transformModel(e))
+      )
     ) as Promise<Model>
   }
 
@@ -252,13 +283,16 @@ export default class JanModelExtension extends ModelExtension {
     option?: OptionType
   ): Promise<void> {
     return this.queue.add(() =>
-      ky
+      this.apiInstance().then((api) =>
-        .post(`${API_URL}/v1/models/import`, {
+        api
-          json: { model, modelPath, name, option },
+          .post('v1/models/import', {
-        })
+            json: { model, modelPath, name, option },
-        .json()
+            timeout: false,
-        .catch((e) => console.debug(e)) // Ignore error
+          })
-        .then()
+          .json()
+          .catch((e) => console.debug(e)) // Ignore error
+          .then()
+      )
     )
   }
 
@@ -269,7 +303,11 @@ export default class JanModelExtension extends ModelExtension {
    */
   async getSources(): Promise<ModelSource[]> {
     const sources = await this.queue
-      .add(() => ky.get(`${API_URL}/v1/models/sources`).json<Data<ModelSource>>())
+      .add(() =>
+        this.apiInstance().then((api) =>
+          api.get('v1/models/sources').json<Data<ModelSource>>()
+        )
+      )
       .then((e) => (typeof e === 'object' ? (e.data as ModelSource[]) : []))
       .catch(() => [])
     return sources.concat(
@@ -283,11 +321,13 @@ export default class JanModelExtension extends ModelExtension {
    */
   async addSource(source: string): Promise<any> {
     return this.queue.add(() =>
-      ky.post(`${API_URL}/v1/models/sources`, {
+      this.apiInstance().then((api) =>
-        json: {
+        api.post('v1/models/sources', {
-          source,
+          json: {
-        },
+            source,
-      })
+          },
+        })
+      )
     )
   }
 
@@ -297,11 +337,14 @@ export default class JanModelExtension extends ModelExtension {
    */
   async deleteSource(source: string): Promise<any> {
     return this.queue.add(() =>
-      ky.delete(`${API_URL}/v1/models/sources`, {
+      this.apiInstance().then((api) =>
-        json: {
+        api.delete('v1/models/sources', {
-          source,
+          json: {
-        },
+            source,
-      })
+          },
+          timeout: false,
+        })
+      )
     )
   }
   // END - Model Sources
@@ -312,7 +355,9 @@ export default class JanModelExtension extends ModelExtension {
    */
   async isModelLoaded(model: string): Promise<boolean> {
     return this.queue
-      .add(() => ky.get(`${API_URL}/v1/models/status/${model}`))
+      .add(() =>
+        this.apiInstance().then((api) => api.get(`v1/models/status/${model}`))
+      )
       .then((e) => true)
       .catch(() => false)
   }
@@ -324,14 +369,18 @@ export default class JanModelExtension extends ModelExtension {
     return this.updateCortexConfig(options).catch((e) => console.debug(e))
   }
 
-   /**
+  /**
    * Fetches models list from cortex.cpp
    * @param model
    * @returns
    */
-   async fetchModels(): Promise<Model[]> {
+  async fetchModels(): Promise<Model[]> {
     return this.queue
-      .add(() => ky.get(`${API_URL}/v1/models?limit=-1`).json<Data<Model>>())
+      .add(() =>
+        this.apiInstance().then((api) =>
+          api.get('v1/models?limit=-1').json<Data<Model>>()
+        )
+      )
       .then((e) =>
         typeof e === 'object' ? e.data.map((e) => this.transformModel(e)) : []
       )
@@ -371,7 +420,9 @@ export default class JanModelExtension extends ModelExtension {
   }): Promise<void> {
     return this.queue
       .add(() =>
-        ky.patch(`${API_URL}/v1/configs`, { json: body }).then(() => {})
+        this.apiInstance().then((api) =>
+          api.patch('v1/configs', { json: body }).then(() => {})
+        )
       )
       .catch((e) => console.debug(e))
   }
@@ -381,14 +432,16 @@ export default class JanModelExtension extends ModelExtension {
    * @returns
    */
   private healthz(): Promise<void> {
-    return ky
+    return this.apiInstance()
-      .get(`${API_URL}/healthz`, {
+      .then((api) =>
-        retry: {
+        api.get('healthz', {
-          limit: 20,
+          retry: {
-          delay: () => 500,
+            limit: 20,
-          methods: ['get'],
+            delay: () => 500,
-        },
+            methods: ['get'],
-      })
+          },
+        })
+      )
       .then(() => {
         this.queue.concurrency = Infinity
       })
@@ -401,17 +454,22 @@ export default class JanModelExtension extends ModelExtension {
     const models = await this.fetchModels()
 
     return this.queue.add(() =>
-      ky
+      this.apiInstance()
-        .get(`${API_URL}/v1/models/hub?author=cortexso&tag=cortex.cpp`)
+        .then((api) =>
-        .json<Data<string>>()
+          api
-        .then((e) => {
+            .get('v1/models/hub?author=cortexso&tag=cortex.cpp')
-          e.data?.forEach((model) => {
+            .json<Data<string>>()
-            if (
+            .then((e) => {
-              !models.some((e) => 'modelSource' in e && e.modelSource === model)
+              e.data?.forEach((model) => {
-            )
+                if (
-              this.addSource(model).catch((e) => console.debug(e))
+                  !models.some(
-          })
+                    (e) => 'modelSource' in e && e.modelSource === model
-        })
+                  )
+                )
+                  this.addSource(model).catch((e) => console.debug(e))
+              })
+            })
+        )
         .catch((e) => console.debug(e))
     )
   }

server/cortex.json

@@ -2668,7 +2668,7 @@
                       },
                       "url": {
                         "type": "string",
-                        "example": "https://api.github.com/repos/janhq/cortex.llamacpp/releases/186479804"
+                        "example": "https://api.github.com/repos/menloresearch/cortex.llamacpp/releases/186479804"
                       }
                     }
                   }
@@ -3633,238 +3633,6 @@
         },
         "tags": ["Files"]
       }
-    },
-    "/configs": {
-      "get": {
-        "summary": "Get Configurations",
-        "description": "Retrieves the current configuration settings of the Cortex server.",
-        "responses": {
-          "200": {
-            "description": "Successful response",
-            "content": {
-              "application/json": {
-                "schema": {
-                  "type": "object",
-                  "properties": {
-                    "allowed_origins": {
-                      "type": "array",
-                      "items": {
-                        "type": "string"
-                      },
-                      "example": ["http://127.0.0.1:39281", "https://cortex.so"]
-                    },
-                    "cors": {
-                      "type": "boolean",
-                      "example": false
-                    },
-                    "proxy_username": {
-                      "type": "string",
-                      "example": "username"
-                    },
-                    "proxy_password": {
-                      "type": "string",
-                      "example": "password"
-                    },
-                    "proxy_url": {
-                      "type": "string",
-                      "example": "http://proxy.example.com:8080"
-                    },
-                    "verify_proxy_ssl": {
-                      "type": "boolean",
-                      "description": "test",
-                      "example": false
-                    },
-                    "verify_proxy_host_ssl": {
-                      "type": "boolean",
-                      "example": false
-                    },
-                    "verify_peer_ssl": {
-                      "type": "boolean",
-                      "example": false
-                    },
-                    "verify_host_ssl": {
-                      "type": "boolean",
-                      "example": false
-                    },
-                    "no_proxy": {
-                      "type": "string",
-                      "example": "localhost"
-                    },
-                    "huggingface_token": {
-                      "type": "string",
-                      "example": "your_token"
-                    }
-                  }
-                },
-                "example": {
-                  "allowed_origins": [
-                    "http://127.0.0.1:39281",
-                    "https://cortex.so"
-                  ],
-                  "cors": false,
-                  "proxy_username": "username",
-                  "proxy_password": "password",
-                  "proxy_url": "http://proxy.example.com:8080",
-                  "verify_proxy_ssl": false,
-                  "verify_proxy_host_ssl": false,
-                  "verify_peer_ssl": false,
-                  "verify_host_ssl": false,
-                  "no_proxy": "localhost",
-                  "huggingface_token": "your_token"
-                }
-              }
-            }
-          }
-        },
-        "tags": ["Configurations"]
-      },
-      "patch": {
-        "tags": ["Configurations"],
-        "summary": "Update configuration settings",
-        "requestBody": {
-          "required": true,
-          "content": {
-            "application/json": {
-              "schema": {
-                "type": "object",
-                "properties": {
-                  "cors": {
-                    "type": "boolean",
-                    "description": "Indicates whether CORS is enabled.",
-                    "example": false
-                  },
-                  "allowed_origins": {
-                    "type": "array",
-                    "items": {
-                      "type": "string"
-                    },
-                    "description": "List of allowed origins.",
-                    "example": ["http://127.0.0.1:39281", "https://cortex.so"]
-                  },
-                  "proxy_username": {
-                    "type": "string",
-                    "description": "Username for the proxy server.",
-                    "example": "username"
-                  },
-                  "proxy_password": {
-                    "type": "string",
-                    "description": "Password for the proxy server.",
-                    "example": "password"
-                  },
-                  "proxy_url": {
-                    "type": "string",
-                    "description": "URL for the proxy server.",
-                    "example": "http://proxy.example.com:8080"
-                  },
-                  "verify_proxy_ssl": {
-                    "type": "boolean",
-                    "description": "Indicates whether to verify the SSL certificate of the proxy server.",
-                    "example": false
-                  },
-                  "verify_proxy_host_ssl": {
-                    "type": "boolean",
-                    "description": "Indicates whether to verify the SSL certificate of the proxy server host.",
-                    "example": false
-                  },
-                  "verify_peer_ssl": {
-                    "type": "boolean",
-                    "description": "Indicates whether to verify the SSL certificate of the peer.",
-                    "example": false
-                  },
-                  "verify_host_ssl": {
-                    "type": "boolean",
-                    "description": "Indicates whether to verify the SSL certificate of the host.",
-                    "example": false
-                  },
-                  "no_proxy": {
-                    "type": "string",
-                    "description": "List of hosts that should not be proxied.",
-                    "example": "localhost"
-                  },
-                  "huggingface_token": {
-                    "type": "string",
-                    "description": "HuggingFace token to pull models.",
-                    "example": "your_token"
-                  }
-                }
-              }
-            }
-          }
-        },
-        "responses": {
-          "200": {
-            "description": "Configuration updated successfully",
-            "content": {
-              "application/json": {
-                "schema": {
-                  "type": "object",
-                  "properties": {
-                    "config": {
-                      "type": "object",
-                      "properties": {
-                        "allowed_origins": {
-                          "type": "array",
-                          "items": {
-                            "type": "string"
-                          },
-                          "example": [
-                            "http://127.0.0.1:39281",
-                            "https://cortex.so"
-                          ]
-                        },
-                        "cors": {
-                          "type": "boolean",
-                          "example": false
-                        },
-                        "proxy_username": {
-                          "type": "string",
-                          "example": "username"
-                        },
-                        "proxy_password": {
-                          "type": "string",
-                          "example": "password"
-                        },
-                        "proxy_url": {
-                          "type": "string",
-                          "example": "http://proxy.example.com:8080"
-                        },
-                        "verify_proxy_ssl": {
-                          "type": "boolean",
-                          "example": false
-                        },
-                        "verify_proxy_host_ssl": {
-                          "type": "boolean",
-                          "example": false
-                        },
-                        "verify_peer_ssl": {
-                          "type": "boolean",
-                          "example": false
-                        },
-                        "verify_host_ssl": {
-                          "type": "boolean",
-                          "example": false
-                        },
-                        "no_proxy": {
-                          "type": "string",
-                          "example": "localhost"
-                        },
-                        "huggingface_token": {
-                          "type": "string",
-                          "example": "your_token"
-                        }
-                      }
-                    },
-                    "message": {
-                      "type": "string",
-                      "example": "Configuration updated successfully"
-                    }
-                  }
-                }
-              }
-            }
-          }
-        }
-      }
     }
   },
   "info": {

server/index.ts

@@ -86,6 +86,14 @@ export const startServer = async (configs?: ServerConfig): Promise<boolean> => {
       },
     })
 
+    const rewriteRequestHeaders = (req: any, headers: any) => {
+      if (req.url.includes('/configs')) return headers
+      return {
+        ...headers,
+        authorization: `Bearer ${process.env.appToken}`, // Add or modify Authorization header
+      }
+    }
+
     // Register Swagger UI
     await server.register(require('@fastify/swagger-ui'), {
       routePrefix: '/',
@@ -102,24 +110,36 @@ export const startServer = async (configs?: ServerConfig): Promise<boolean> => {
       upstream: `${CORTEX_API_URL}/v1`,
       prefix: configs?.prefix ?? '/v1',
       http2: false,
-    })
+      replyOptions: {
-
+        rewriteRequestHeaders,
-    server.register(proxy, {
+      },
-      upstream: `${CORTEX_API_URL}/system`,
-      prefix:'/system',
-      http2: false,
     })
 
     server.register(proxy, {
       upstream: `${CORTEX_API_URL}/processManager`,
-      prefix:'/processManager',
+      prefix: '/processManager',
       http2: false,
+      replyOptions: {
+        rewriteRequestHeaders,
+      },
+    })
+
+    server.register(proxy, {
+      upstream: `${CORTEX_API_URL}/system`,
+      prefix: '/system',
+      http2: false,
+      replyOptions: {
+        rewriteRequestHeaders,
+      },
     })
 
     server.register(proxy, {
       upstream: `${CORTEX_API_URL}/healthz`,
-      prefix:'/healthz',
+      prefix: '/healthz',
       http2: false,
+      replyOptions: {
+        rewriteRequestHeaders,
+      },
     })
 
     // Start listening for requests

web/containers/Providers/DataLoader.tsx

@@ -15,6 +15,7 @@ import { useDebouncedCallback } from 'use-debounce'
 import useAssistants from '@/hooks/useAssistants'
 import { useGetEngines } from '@/hooks/useEngineManagement'
 import useGetSystemResources from '@/hooks/useGetSystemResources'
+import { useGetHardwareInfo } from '@/hooks/useHardwareManagement'
 import useModels from '@/hooks/useModels'
 import useThreads from '@/hooks/useThreads'
 
@@ -34,6 +35,7 @@ const DataLoader: React.FC = () => {
   const setJanSettingScreen = useSetAtom(janSettingScreenAtom)
   const { getData: loadModels } = useModels()
   const { mutate } = useGetEngines()
+  const { mutate: getHardwareInfo } = useGetHardwareInfo(false)
 
   useThreads()
   useAssistants()
@@ -42,6 +44,7 @@ const DataLoader: React.FC = () => {
   useEffect(() => {
     // Load data once
     loadModels()
+    getHardwareInfo()
     // eslint-disable-next-line react-hooks/exhaustive-deps
   }, [])
   const reloadData = useDebouncedCallback(() => {

web/hooks/useGetLatestRelease.ts

@@ -2,7 +2,9 @@
 import useSWR from 'swr'
 
 const fetchLatestRelease = async (includeBeta: boolean) => {
-  const res = await fetch('https://api.github.com/repos/janhq/jan/releases')
+  const res = await fetch(
+    'https://api.github.com/repos/menloresearch/jan/releases'
+  )
   if (!res.ok) throw new Error('Failed to fetch releases')
 
   const releases = await res.json()

web/hooks/useHardwareManagement.ts

@@ -32,7 +32,7 @@ const getExtension = () =>
 /**
  * @returns A Promise that resolves to an object of list engines.
  */
-export function useGetHardwareInfo() {
+export function useGetHardwareInfo(updatePeriodically: boolean = true) {
   const setCpuUsage = useSetAtom(cpuUsageAtom)
   const setUsedRam = useSetAtom(usedRamAtom)
   const setTotalRam = useSetAtom(totalRamAtom)
@@ -56,7 +56,7 @@ export function useGetHardwareInfo() {
     {
       revalidateOnFocus: false,
       revalidateOnReconnect: false,
-      refreshInterval: 2000,
+      refreshInterval: updatePeriodically ? 2000 : undefined,
     }
   )
 

web/screens/Hub/index.tsx

@@ -80,7 +80,7 @@ const filterOptions = [
   },
 ]
 
-const hubCompatibleAtom = atom(true)
+const hubCompatibleAtom = atom(false)
 
 const HubScreen = () => {
   const { sources } = useGetModelSources()

web/screens/Settings/Hardware/index.tsx

@@ -6,13 +6,14 @@ import { useState } from 'react'
 import { DragDropContext, Draggable, Droppable } from '@hello-pangea/dnd'
 
 import { Progress, ScrollArea, Switch } from '@janhq/joi'
-import { useAtom, useAtomValue } from 'jotai'
+import { useAtom, useAtomValue, useSetAtom } from 'jotai'
 import { atomWithStorage } from 'jotai/utils'
 
 import { ChevronDownIcon, GripVerticalIcon } from 'lucide-react'
 
 import { twMerge } from 'tailwind-merge'
 
+import { activeModelAtom } from '@/hooks/useActiveModel'
 import {
   useGetHardwareInfo,
   setActiveGpus,
@@ -47,6 +48,7 @@ const Hardware = () => {
   const ramUtilitized = useAtomValue(ramUtilitizedAtom)
   const showScrollBar = useAtomValue(showScrollBarAtom)
   const [gpus, setGpus] = useAtom(gpusAtom)
+  const setActiveModel = useSetAtom(activeModelAtom)
 
   const [orderGpus, setOrderGpus] = useAtom(orderGpusAtom)
 
@@ -70,11 +72,15 @@ const Hardware = () => {
         .filter((gpu: any) => gpu.activated)
         .map((gpu: any) => Number(gpu.id))
       await setActiveGpus({ gpus: activeGpuIds })
+      setActiveModel(undefined)
       mutate()
-      window.location.reload()
     } catch (error) {
       console.error('Failed to update active GPUs:', error)
     }
+    setIsActivatingGpu((prev) => {
+      prev.delete(id)
+      return new Set(prev)
+    })
   }
 
   const handleDragEnd = (result: any) => {