Perplexity

The finance_search tool is now available in the Agent API. Pull structured financial and market data — quotes, financials, earnings, analyst estimates, segment KPIs, ETF constituents, and more — for public companies and instruments. The model decides which fields to fetch based on your prompt.

Highlights:

• Quotes and pricing: Near-real-time prices, OHLCV ranges, pre-market and after-hours data
• Financials: Income statement, balance sheet, cash flow (quarterly and annual), key ratios
• Earnings: Last call transcript, filings, beat/miss history, guidance
• Coverage and market activity: Analyst estimates, top gainers/losers, ownership and corporate actions
• Recommended configurations: Presets for live quotes, single-company historical lookups, and multi-step cross-company research

Agent API: New Third-Party Models — Claude Opus 4.7, GPT-5.5, and Grok 4.20 Reasoning now supported.

API Key Management: Security Upgrade — Full token values now returned only at creation and cannot be retrieved again from the console or any endpoint. Always set a descriptive token_name.

New Integration: n8n — Native Perplexity node with full API coverage (Chat Completions, Agent, Search, Embeddings). Models load dynamically.

New Integration: OpenClaw — Open-source terminal AI agent now supports Perplexity Search API as a native web search provider.

API Credits via AWS Marketplace — Purchase API credits through your AWS account for consolidated billing.

/v1/models Endpoint — New GET /v1/models lists all available Agent API models in OpenAI-compatible format. No authentication required.

Agent API: New Third-Party Models — GPT-5.4, NVIDIA Nemotron, Claude Sonnet 4.6, and Gemini 3.1 Pro Preview now supported.

Model Deprecations — google/gemini-2.5-flash deprecated March 20, 2026; google/gemini-2.5-pro and google/gemini-3-pro-preview deprecated April 1, 2026.

Agent API Endpoint /v1/agent — Canonical endpoint is now /v1/agent. The previous /v1/responses path continues to work as an alias for OpenAI compatibility.

The Agent API is now generally available. Build autonomous agents with production-ready guidance on model behavior, output controls, and OpenAI-compatibility patterns.

The Embeddings API is now generally available with guides for standard and contextualized embeddings, plus best practices for semantic search and retrieval workflows.

Model Deprecation: sonar-reasoning Removed — As of December 15, 2025, sonar-reasoning has been deprecated. Migrate to sonar-reasoning-pro.

New: Media Classifier — Automatically detects when queries would benefit from visual content and includes relevant images or videos. Available exclusively with sonar-pro. Configurable control to enable/disable and override media types.

Search API Enhancements:

• New max_tokens parameter: Control maximum tokens extracted per page
• last_updated_filter support: Filter results by when content was last updated
• Vercel AI SDK Support: Search API now compatible with the Vercel AI SDK

Community Showcase: Perplexity Client — Electron-based desktop application with advanced API parameter controls.

The Perplexity MCP Server now supports one-click installation for Cursor, VS Code, Claude Desktop, and Claude Code. Provides four tools: perplexity_search, perplexity_ask, perplexity_research, and perplexity_reason.

Pro Search is now generally available for Sonar Pro. Enhances queries with automated tool usage and multi-step reasoning through intelligent tool orchestration.

Key capabilities:

• Multi-step reasoning: Automatic multiple web searches and URL content fetching
• Real-time thought streaming
• Automatic classification via search_type: "auto"
• Built-in tools: web_search and fetch_url_content

• language_preference: Specify preferred languages for search results (available for sonar and sonar-pro)
• search_domain_filter: Filter results to specific domains
• Date/time filters: Enhanced control over result freshness with publication and update filters

Official Perplexity SDKs for Python and TypeScript are now available.

Installation:

# Python
pip install perplexityai

# TypeScript
npm install @perplexity-ai/perplexity_ai

Features:

• Full type definitions for all request/response fields
• Support for Sonar and Search APIs
• Streaming support with async iterators
• Automatic environment variable handling for API keys

New interactive playground to test Search API queries and parameters in real time — no API key required to get started. Experiment with filtering options, see response structures, and refine queries before implementing in code.

Standalone Search API providing direct access to search results without LLM processing. Returns raw, ranked search results from Perplexity's continuously refreshed index.

Perfect for: building custom search experiences, integrating raw search results into applications, specialized workflows.

Key features:

• Direct access to Perplexity's search index
• All existing search filters and controls
• Faster responses (no LLM processing)
• Same filtering options (domain, date range, academic sources, etc.)

Upload and analyze documents in multiple formats using Sonar models. Supports PDF, DOC, DOCX, TXT, and RTF files.

Key capabilities:

• Document Analysis: Ask questions about document content
• Content Extraction: Pull out key information and insights
• Multi-format Support: PDF, Word, text, and Rich Text Format
• Large Document Handling and Multi-language Support

Upload via publicly accessible URLs using the file_url content type.

As of August 1, 2025, R1-1776 has been removed from the available models. Recommended replacement: Sonar Pro Reasoning.

New SEC filings filter allows searching specifically within SEC regulatory documents. Set search_domain: "sec" to focus on 10-K, 10-Q, 8-K, and other regulatory filings.

Works with other search parameters like date filters and search context size. Useful for financial analysts, investment professionals, and compliance officers.

API response JSON now includes detailed cost information for each request via a new cost object in the usage field.

"usage": {
  "prompt_tokens": 8,
  "completion_tokens": 439,
  "total_tokens": 447,
  "search_context_size": "low",
  "cost": {
    "input_tokens_cost": 2.4e-05,
    "output_tokens_cost": 0.006585,
    "request_cost": 0.006,
    "total_cost": 0.012609
  }
}

Includes: input_tokens_cost, output_tokens_cost, request_cost, total_cost.

New academic filter allows tailoring searches to academic and scholarly sources. Set search_mode: "academic" to prioritize results from peer-reviewed papers, journal articles, and research publications.

Works with search_context_size and date filters. Ideal for students, researchers, and professionals requiring scientifically accurate information.

New latest_updated field added to date range filtering, allowing filtering based on when a webpage was last modified.

New fields:

• latest_updated: Filter by last modification date
• published_after / published_before: Existing publication date filters

Useful for finding the most current version of frequently updated content.

New search_results field added to API response JSON, providing direct access to search results used by models.

Fields included:

• title: Title of the search result page
• url: URL of the search result
• date: Publication date of the content

Breaking Change: The citations field has been fully deprecated and removed. All applications must now use the search_results field.

New reasoning_effort parameter for sonar-deep-research to control computational effort per query.

Options:

• "low": Faster, simpler answers with reduced token usage
• "medium": Balanced approach (default)
• "high": Deeper, more thorough responses with increased token usage

Directly impacts reasoning tokens consumed, enabling cost control.