Remove Embeddings from LLM feature

apps/client/src/server_types.ts

@@ -23,4 +23,4 @@ export interface EntityChange {
     instanceId?: string | null;
 }
 
-export type EntityType = "notes" | "branches" | "attributes" | "note_reordering" | "revisions" | "options" | "attachments" | "blobs" | "etapi_tokens" | "note_embeddings";
+export type EntityType = "notes" | "branches" | "attributes" | "note_reordering" | "revisions" | "options" | "attachments" | "blobs" | "etapi_tokens";

apps/client/src/services/froca_updater.ts

@@ -35,7 +35,7 @@ async function processEntityChanges(entityChanges: EntityChange[]) {
                 loadResults.addOption(attributeEntity.name);
             } else if (ec.entityName === "attachments") {
                 processAttachment(loadResults, ec);
-            } else if (ec.entityName === "blobs" || ec.entityName === "etapi_tokens" || ec.entityName === "note_embeddings") {
+            } else if (ec.entityName === "blobs" || ec.entityName === "etapi_tokens") {
                 // NOOP - these entities are handled at the backend level and don't require frontend processing
             } else {
                 throw new Error(`Unknown entityName '${ec.entityName}'`);

apps/client/src/services/load_results.ts

@@ -44,18 +44,7 @@ interface OptionRow {}
 
 interface NoteReorderingRow {}
 
-interface NoteEmbeddingRow {
-    embedId: string;
-    noteId: string;
-    providerId: string;
-    modelId: string;
-    dimension: number;
-    version: number;
-    dateCreated: string;
-    utcDateCreated: string;
-    dateModified: string;
-    utcDateModified: string;
-}
+
 
 type EntityRowMappings = {
     notes: NoteRow;
@@ -64,7 +53,6 @@ type EntityRowMappings = {
     options: OptionRow;
     revisions: RevisionRow;
     note_reordering: NoteReorderingRow;
-    note_embeddings: NoteEmbeddingRow;
 };
 
 export type EntityRowNames = keyof EntityRowMappings;

apps/client/src/translations/en/translation.json

@@ -1124,18 +1124,15 @@
     "layout-horizontal-description": "launcher bar is underneath the tab bar, the tab bar is now full width."
   },
   "ai_llm": {
-    "embeddings_configuration": "Embeddings Configuration",
     "not_started": "Not started",
-    "title": "AI & Embedding Settings",
-    "embedding_statistics": "Embedding Statistics",
+    "title": "AI Settings",
     "processed_notes": "Processed Notes",
     "total_notes": "Total Notes",
     "progress": "Progress",
     "queued_notes": "Queued Notes",
     "failed_notes": "Failed Notes",
     "last_processed": "Last Processed",
     "refresh_stats": "Refresh Statistics",
-    "no_failed_embeddings": "No failed embeddings found.",
     "enable_ai_features": "Enable AI/LLM features",
     "enable_ai_description": "Enable AI features like note summarization, content generation, and other LLM capabilities",
     "openai_tab": "OpenAI",
@@ -1160,49 +1157,27 @@
     "anthropic_api_key_description": "Your Anthropic API key for accessing Claude models",
     "default_model": "Default Model",
     "openai_model_description": "Examples: gpt-4o, gpt-4-turbo, gpt-3.5-turbo",
-    "embedding_model": "Embedding Model",
-    "openai_embedding_model_description": "Model used for generating embeddings (text-embedding-3-small recommended)",
     "base_url": "Base URL",
     "openai_url_description": "Default: https://api.openai.com/v1",
     "anthropic_settings": "Anthropic Settings",
     "anthropic_url_description": "Base URL for the Anthropic API (default: https://api.anthropic.com)",
     "anthropic_model_description": "Anthropic Claude models for chat completion",
     "voyage_settings": "Voyage AI Settings",
-    "voyage_api_key_description": "Your Voyage AI API key for accessing embeddings services",
     "ollama_settings": "Ollama Settings",
     "ollama_url_description": "URL for the Ollama API (default: http://localhost:11434)",
     "ollama_model_description": "Ollama model to use for chat completion",
     "anthropic_configuration": "Anthropic Configuration",
-    "voyage_embedding_model_description": "Voyage AI embedding models for text embeddings (voyage-2 recommended)",
     "voyage_configuration": "Voyage AI Configuration",
     "voyage_url_description": "Default: https://api.voyageai.com/v1",
     "ollama_configuration": "Ollama Configuration",
     "enable_ollama": "Enable Ollama",
     "enable_ollama_description": "Enable Ollama for local AI model usage",
     "ollama_url": "Ollama URL",
     "ollama_model": "Ollama Model",
-    "ollama_embedding_model": "Embedding Model",
-    "ollama_embedding_model_description": "Specialized model for generating embeddings (vector representations)",
     "refresh_models": "Refresh Models",
     "refreshing_models": "Refreshing...",
-    "embedding_configuration": "Embeddings Configuration",
-    "embedding_default_provider": "Default Provider",
-    "embedding_default_provider_description": "Select the default provider used for generating note embeddings",
-    "embedding_provider_precedence": "Embedding Provider Precedence",
-    "embedding_providers_order": "Embedding Provider Order",
-    "embedding_providers_order_description": "Set the order of embedding providers in comma-separated format (e.g., \"openai,voyage,ollama,local\")",
     "enable_automatic_indexing": "Enable Automatic Indexing",
-    "enable_automatic_indexing_description": "Automatically generate embeddings for new and updated notes",
-    "embedding_auto_update_enabled": "Auto-update Embeddings",
-    "embedding_auto_update_enabled_description": "Automatically update embeddings when notes are modified",
-    "recreate_embeddings": "Recreate All Embeddings",
-    "recreate_embeddings_description": "Regenerate all note embeddings from scratch (may take a long time for large note collections)",
-    "recreate_embeddings_started": "Embeddings regeneration started. This may take a long time for large note collections.",
-    "recreate_embeddings_error": "Error starting embeddings regeneration. Check logs for details.",
-    "recreate_embeddings_confirm": "Are you sure you want to recreate all embeddings? This may take a long time for large note collections.",
     "rebuild_index": "Rebuild Index",
-    "rebuild_index_description": "Rebuild the vector search index for better performance (much faster than recreating embeddings)",
-    "rebuild_index_started": "Embedding index rebuild started. This may take several minutes.",
     "rebuild_index_error": "Error starting index rebuild. Check logs for details.",
     "note_title": "Note Title",
     "error": "Error",
@@ -1212,43 +1187,16 @@
     "partial": "{{ percentage }}% completed",
     "retry_queued": "Note queued for retry",
     "retry_failed": "Failed to queue note for retry",
-    "embedding_provider_precedence_description": "Comma-separated list of providers in order of precedence for embeddings search (e.g., 'openai,ollama,anthropic')",
-    "embedding_dimension_strategy": "Embedding Dimension Strategy",
-    "embedding_dimension_auto": "Auto (Recommended)",
-    "embedding_dimension_fixed": "Fixed",
-    "embedding_similarity_threshold": "Similarity Threshold",
-    "embedding_similarity_threshold_description": "Minimum similarity score for notes to be included in search results (0-1)",
     "max_notes_per_llm_query": "Max Notes Per Query",
     "max_notes_per_llm_query_description": "Maximum number of similar notes to include in AI context",
-    "embedding_dimension_strategy_description": "Choose how embeddings are handled. 'Native' preserves maximum information by adapting smaller vectors to match larger ones (recommended). 'Regenerate' creates new embeddings with the target model for specific search needs.",
-    "drag_providers_to_reorder": "Drag providers up or down to set your preferred order for embedding searches",
     "active_providers": "Active Providers",
     "disabled_providers": "Disabled Providers",
     "remove_provider": "Remove provider from search",
     "restore_provider": "Restore provider to search",
-    "embedding_generation_location": "Generation Location",
-    "embedding_generation_location_description": "Select where embedding generation should happen",
-    "embedding_generation_location_client": "Client/Server",
-    "embedding_generation_location_sync_server": "Sync Server",
-    "enable_auto_update_embeddings": "Auto-update Embeddings",
-    "enable_auto_update_embeddings_description": "Automatically update embeddings when notes are modified",
-    "auto_update_embeddings": "Auto-update Embeddings",
-    "auto_update_embeddings_desc": "Automatically update embeddings when notes are modified",
     "similarity_threshold": "Similarity Threshold",
     "similarity_threshold_description": "Minimum similarity score (0-1) for notes to be included in context for LLM queries",
-    "embedding_batch_size": "Batch Size",
-    "embedding_batch_size_description": "Number of notes to process in a single batch (1-50)",
-    "embedding_update_interval": "Update Interval (ms)",
-    "embedding_update_interval_description": "Time between processing batches of embeddings (in milliseconds)",
-    "embedding_default_dimension": "Default Dimension",
-    "embedding_default_dimension_description": "Default embedding vector dimension when creating new embeddings",
-    "reprocess_all_embeddings": "Reprocess All Embeddings",
-    "reprocess_all_embeddings_description": "Queue all notes for embedding processing. This may take some time depending on your number of notes.",
-    "reprocessing_embeddings": "Reprocessing...",
-    "reprocess_started": "Embedding reprocessing started in the background",
-    "reprocess_error": "Error starting embedding reprocessing",
+
     "reprocess_index": "Rebuild Search Index",
-    "reprocess_index_description": "Optimize the search index for better performance. This uses existing embeddings without regenerating them (much faster than reprocessing all embeddings).",
     "reprocessing_index": "Rebuilding...",
     "reprocess_index_started": "Search index optimization started in the background",
     "reprocess_index_error": "Error rebuilding search index",
@@ -1261,15 +1209,13 @@
     "incomplete": "Incomplete ({{percentage}}%)",
     "complete": "Complete (100%)",
     "refreshing": "Refreshing...",
-    "stats_error": "Error fetching embedding statistics",
     "auto_refresh_notice": "Auto-refreshes every {{seconds}} seconds",
     "note_queued_for_retry": "Note queued for retry",
     "failed_to_retry_note": "Failed to retry note",
     "all_notes_queued_for_retry": "All failed notes queued for retry",
     "failed_to_retry_all": "Failed to retry notes",
     "ai_settings": "AI Settings",
     "api_key_tooltip": "API key for accessing the service",
-    "confirm_delete_embeddings": "Are you sure you want to delete all AI embeddings? This will remove all semantic search capabilities until notes are reindexed, which can take a significant amount of time.",
     "empty_key_warning": {
       "anthropic": "Anthropic API key is empty. Please enter a valid API key.",
       "openai": "OpenAI API key is empty. Please enter a valid API key.",
@@ -1302,7 +1248,6 @@
     "note_chat": "Note Chat",
     "notes_indexed": "{{ count }} note indexed",
     "notes_indexed_plural": "{{ count }} notes indexed",
-    "reset_embeddings": "Reset Embeddings",
     "sources": "Sources",
     "start_indexing": "Start Indexing",
     "use_advanced_context": "Use Advanced Context",
@@ -1315,24 +1260,10 @@
     },
     "create_new_ai_chat": "Create new AI Chat",
     "configuration_warnings": "There are some issues with your AI configuration. Please check your settings.",
-    "embeddings_started": "Embedding generation started",
-    "embeddings_stopped": "Embedding generation stopped",
-    "embeddings_toggle_error": "Error toggling embeddings",
-    "local_embedding_description": "Uses local embedding models for offline text embedding generation",
-    "local_embedding_settings": "Local Embedding Settings",
-    "ollama_embedding_settings": "Ollama Embedding Settings",
-    "ollama_embedding_url_description": "URL for the Ollama API for embedding generation (default: http://localhost:11434)",
-    "openai_embedding_api_key_description": "Your OpenAI API key for embedding generation (can be different from chat API key)",
-    "openai_embedding_settings": "OpenAI Embedding Settings",
-    "openai_embedding_url_description": "Base URL for OpenAI embedding API (default: https://api.openai.com/v1)",
-    "selected_embedding_provider": "Selected Embedding Provider",
-    "selected_embedding_provider_description": "Choose the provider for generating note embeddings",
     "selected_provider": "Selected Provider",
     "selected_provider_description": "Choose the AI provider for chat and completion features",
-    "select_embedding_provider": "Select embedding provider...",
     "select_model": "Select model...",
-    "select_provider": "Select provider...",
-    "voyage_embedding_url_description": "Base URL for the Voyage AI embedding API (default: https://api.voyageai.com/v1)"
+    "select_provider": "Select provider..."
   },
   "zoom_factor": {
     "title": "Zoom Factor (desktop build only)",

apps/client/src/widgets/llm_chat/communication.ts

@@ -258,9 +258,3 @@ export async function getDirectResponse(noteId: string, messageParams: any): Pro
     }
 }
 
-/**
- * Get embedding statistics
- */
-export async function getEmbeddingStats(): Promise<any> {
-    return server.get('llm/embeddings/stats');
-}

apps/client/src/widgets/llm_chat/llm_chat_panel.ts

@@ -11,7 +11,7 @@ import { TPL, addMessageToChat, showSources, hideSources, showLoadingIndicator,
 import { formatMarkdown } from "./utils.js";
 import { createChatSession, checkSessionExists, setupStreamingResponse, getDirectResponse } from "./communication.js";
 import { extractInChatToolSteps } from "./message_processor.js";
-import { validateEmbeddingProviders } from "./validation.js";
+import { validateProviders } from "./validation.js";
 import type { MessageData, ToolExecutionStep, ChatData } from "./types.js";
 import { formatCodeBlocks } from "../../services/syntax_highlight.js";
 import { ClassicEditor, type CKTextEditor, type MentionFeed } from "@triliumnext/ckeditor5";
@@ -616,7 +616,7 @@ export default class LlmChatPanel extends BasicWidget {
         }
 
         // Check for any provider validation issues when refreshing
-        await validateEmbeddingProviders(this.validationWarning);
+        await validateProviders(this.validationWarning);
 
         // Get current note context if needed
         const currentActiveNoteId = appContext.tabManager.getActiveContext()?.note?.noteId || null;
@@ -767,7 +767,7 @@ export default class LlmChatPanel extends BasicWidget {
      */
     private async processUserMessage(content: string) {
         // Check for validation issues first
-        await validateEmbeddingProviders(this.validationWarning);
+        await validateProviders(this.validationWarning);
 
         // Make sure we have a valid session
         if (!this.noteId) {

apps/client/src/widgets/llm_chat/validation.ts

@@ -2,12 +2,11 @@
  * Validation functions for LLM Chat
  */
 import options from "../../services/options.js";
-import { getEmbeddingStats } from "./communication.js";
 
 /**
- * Validate embedding providers configuration
+ * Validate providers configuration
  */
-export async function validateEmbeddingProviders(validationWarning: HTMLElement): Promise<void> {
+export async function validateProviders(validationWarning: HTMLElement): Promise<void> {
     try {
         // Check if AI is enabled
         const aiEnabled = options.is('aiEnabled');
@@ -62,23 +61,8 @@ export async function validateEmbeddingProviders(validationWarning: HTMLElement)
             // Add checks for other providers as needed
         }
 
-        // Fetch embedding stats to check if there are any notes being processed
-        const embeddingStats = await getEmbeddingStats() as {
-            success: boolean,
-            stats: {
-                totalNotesCount: number;
-                embeddedNotesCount: number;
-                queuedNotesCount: number;
-                failedNotesCount: number;
-                lastProcessedDate: string | null;
-                percentComplete: number;
-            }
-        };
-        const queuedNotes = embeddingStats?.stats?.queuedNotesCount || 0;
-        const hasEmbeddingsInQueue = queuedNotes > 0;
-
-        // Show warning if there are configuration issues or embeddings in queue
-        if (configIssues.length > 0 || hasEmbeddingsInQueue) {
+        // Show warning if there are configuration issues
+        if (configIssues.length > 0) {
             let message = '<i class="bx bx-error-circle me-2"></i><strong>AI Provider Configuration Issues</strong>';
 
             message += '<ul class="mb-1 ps-4">';
@@ -87,11 +71,6 @@ export async function validateEmbeddingProviders(validationWarning: HTMLElement)
             for (const issue of configIssues) {
                 message += `<li>${issue}</li>`;
             }
-
-            // Show warning about embeddings queue if applicable
-            if (hasEmbeddingsInQueue) {
-                message += `<li>Currently processing embeddings for ${queuedNotes} notes. Some AI features may produce incomplete results until processing completes.</li>`;
-            }
 
             message += '</ul>';
             message += '<div class="mt-2"><a href="javascript:" class="settings-link btn btn-sm btn-outline-secondary"><i class="bx bx-cog me-1"></i>Open AI Settings</a></div>';
@@ -103,7 +82,7 @@ export async function validateEmbeddingProviders(validationWarning: HTMLElement)
             validationWarning.style.display = 'none';
         }
     } catch (error) {
-        console.error('Error validating embedding providers:', error);
+        console.error('Error validating providers:', error);
         validationWarning.style.display = 'none';
     }
 }