--- title: "Unified search" method: POST path: "/search" tags: ["Search"] --- # Unified search `POST /search` Single search endpoint with a mode discriminator. ## Request body - union - QuerySearchRequest - `messages` object[], required — List of chat messages - `repositories` union[] — List of repositories to query. Can be strings (slug, display name) or dicts with a 'repository' field. - union - string - object - `data_sources` union[] — List of data sources to query. Can be strings (display_name, URL, or source_id) or dicts with 'source_id' or 'identifier' fields - union - string - object - `local_folders` union[] — List of local folders to query. Can be strings (display_name or local_folder_id) or dicts with 'local_folder_id' or 'identifier' fields. Local folders are private and user-scoped. - union - string - object - `slack_workspaces` string[] — List of Slack installation IDs to include in search - `slack_filters` SlackSearchFilters — Filters for Slack message search within query endpoint. - `channels` string[], nullable — Filter by channel names or IDs - `users` string[], nullable — Filter by Slack user names or IDs - `date_from` string, nullable — Oldest message timestamp (ISO 8601) - `date_to` string, nullable — Newest message timestamp (ISO 8601) - `include_threads` boolean — Include threaded messages in results - `category` string, nullable — Filter local folder results by classification category (e.g., 'Work', 'Personal') - `local_source_filters` LocalSourceFilters — Filters for local/personal source retrieval (messages, contacts, etc.). - `source_subtype` string, nullable — Filter by source subtype (e.g., 'database', 'telegram_export') - `db_type` string, nullable — Filter by database type (e.g., 'imessage', 'telegram', 'safari_history') - `connector_type` string, nullable — Filter by connector type (e.g., 'imessage', 'whatsapp', 'contacts') - `conversation_id` string, nullable — Filter to a specific conversation/chat - `contact_id` string, nullable — Filter to messages with a specific contact - `sender_role` string, nullable — Filter by sender role: 'self' or 'contact' - `time_after` string, nullable — Only return results with timestamp after this ISO datetime - `time_before` string, nullable — Only return results with timestamp before this ISO datetime - `source_trust_filter` SourceTrustFilter — Filtering preferences for trust-aware retrieval. - `minimum_trust_tier` 'low' | 'medium' | 'high' | 'verified', nullable — Only include sources at or above this effective trust tier - `verified_only` boolean — Only include Nia-verified sources - `require_overlay` boolean — Only include sources with curated overlays - `search_mode` string — Search mode: 'repositories', 'sources', or 'unified' - `stream` boolean — Whether to stream the response - `include_sources` boolean — Whether to include source texts in the response - `fast_mode` boolean — Skip LLM processing for faster results (100-500ms vs 2-8s). Set to false for deeper analysis. - `skip_llm` boolean — Return raw search results without any LLM processing. Returns only sources with scores. - `reasoning_strategy` string — Retrieval strategy: 'vector' (default similarity search), 'tree' (LLM-guided tree navigation for PDFs), or 'hybrid' (both combined) - `max_tokens` integer, nullable — Maximum tokens in response. Results truncated when budget reached. - `model` string, nullable — Synthesis model override. Allowed: claude-sonnet-4-5, gpt-5.2-2025-12-11, claude-haiku-4-5-20251001 - `semantic_cache_threshold` number — Minimum similarity for semantic cache hit (non-streaming only) - `bypass_semantic_cache` boolean — Skip semantic cache (L2) lookup - `include_follow_ups` boolean — Generate follow-up questions (adds ~500-1000ms LLM latency). Disabled by default for speed. - `e2e_session_id` string, nullable — Active E2E decrypt session ID. When provided, encrypted local folder results are decrypted through the desktop bridge before synthesis. - `mode` 'query' — Search mode discriminator - WebSearchRequestWithMode - `query` string, required — Search query - `num_results` integer — Number of results - `category` string, nullable — Filter by category: github, company, research, news, tweet, pdf, blog - `days_back` integer, nullable — Only show results from last N days - `find_similar_to` string, nullable — URL to find similar content to - `mode` 'web' — Search mode discriminator - DeepResearchRequestWithMode - `query` string, required — Research question - `output_format` string, nullable — Optional structure hint - `verbose` boolean — Include verbose trace output - `mode` 'deep' — Search mode discriminator - UniversalSearchRequestWithMode - `query` string, required — Natural language search query - `top_k` integer — Total number of results - `include_repos` boolean — Include repository sources - `include_docs` boolean — Include documentation sources - `include_huggingface_datasets` boolean — Include HuggingFace dataset sources (excluded by default to prevent search pollution) - `alpha` number — Weight for vector search (0.7 = 70% vector) - `compress_output` boolean — Use AI to compress results - `max_sources` integer — Max source namespaces to deep search - `sources_for_answer` integer — Number of results to use for AI answer - `bypass_cache` boolean — Skip cache and force fresh search results - `max_tokens` integer, nullable — Maximum tokens in response. Results truncated when budget reached. - `boost_source_types` object — Source type boosts (override to customize or set {} to disable) - `boost_languages` string[], nullable — Programming languages to boost (e.g., ['python', 'typescript']) - `language_boost_factor` number — Boost multiplier for preferred languages - `use_native_boosting` boolean — Use TurboPuffer FTS v2 native Sum/Product boosting - `semantic_cache_threshold` number — Minimum similarity for semantic cache hit (0.8-1.0) - `bypass_semantic_cache` boolean — Skip semantic cache (L2) lookup - `expand_symbols` boolean — Extract function/class names from results and search for usages (cAST-inspired) - `mode` 'universal' — Search mode discriminator ## Response `200` Successful Response - unknown ## Other responses - `422` — Validation Error --- [API](https://skmtc.net/nozomio-labs/apis/nia-ai-api.md) · [All operations](https://skmtc.net/nozomio-labs/apis/nia-ai-api/llms.txt) · [OpenAPI document](https://skmtc-service-staging.skmtc.workers.dev/v1/apis/nozomio-labs/nia-ai-api/revisions/dbb602a3ef17/schema)