Cycle on Jun 1, 2026

I will start by listing the open threads of interest for this research topic so we can align our cycle with the user-prioritized questions.

I will perform a web search to understand the positioning, features, technology, data sources, and business models of Gemini Deep Research, Perplexity (including its deep research/reasoning modes), and Sacra, as well as how Google Scholar fits into this landscape or is being augmented/challenged by AI academic research tools (like Elicit, Consensus, SciSpace).

I will fetch some of the key pages describing Gemini Deep Research, specifically the Google Blog post and the Section AI comparison to understand features, pricing, and how it works.

I will search for OpenAI's "Operator" or "Deep Research" as well as Perplexity's Deep Research or Pro Search, to see how the market is shaping up. Let's do a web search for Perplexity's deep research features, pricing, and technology. I should also search for OpenAI Deep Research. Let's start with Perplexity.

Let's fetch the official Perplexity Deep Research announcement from `perplexity.ai/hub/blog/introducing-perplexity-deep-research`. This will give us concrete details on features, technology, data sources, pricing/business model, and target audience.

Now I will search for Sacra. The user specifically mentioned Sacra as "another research platform." Let's search for what Sacra is, what it does, its features, value prop, data sources, business model, and target audience. Let's do a web search.

Let's search Sacra's website or other sources to understand what they do. Let's run a search for `site:sacra.com` to understand their product lines, like Sacra Pipeline, Sacra Signals, and their research reports. Let's see what Sacra is about.

Let's search for "Sacra" business model, founders, pricing, and how it uses AI. Is it an AI research tool? Or is it a research platform of human equity research on private markets that uses AI to ingest data or generate "Sacra Signals"? Let's search for `site:sacra.com` and its core messaging. Let's do a web search.

Let's search for Sacra's pricing page or subscription cost. Let's do a search for `site:sacra.com/pricing` or `site:sacra.com/subscribe` or `Sacra subscription cost` or `Sacra pricing`.

Let's fetch `https://sacra.com/pricing/` and `https://sacra.com/features/` or `https://sacra.com/p/introducing-sacra-pipeline/` to get a deep understanding of Sacra. Let's fetch the pricing page first.

This is very interesting! Sacra has changed. It is no longer just a static research newsletter. It offers "tasks", "Model Context Protocol (MCP) server access", "API", "Bespoke research", and "Revenue & valuation models". Let's search to understand Sacra's new feature set, "tasks", and how it works as an AI-enabled or agent-ready platform. Let's do a web search.

This is absolutely fascinating. Let's fetch the Sacra MCP / API docs and features to get a very clear picture of what they are doing. Let's do a web fetch on `https://sacra.com/features/mcp/` and `https://docs.sacra.com/` (or search for it if we can't fetch it directly, or try to fetch `https://sacra.com/p/introducing-sacra-mcp-connector-chatgpt/`). Let's fetch `https://sacra.com/features/mcp/` first.

Let's fetch `https://sacra.com/p/introducing-sacra-mcp-connector-chatgpt/` or search for its content to see how it works. Let's do a search first to see if it exists and what the content is. Or we can web_fetch `https://sacra.com/p/introducing-sacra-mcp-connector-chatgpt/`. Let's try to fetch it.

This is highly informative. Now we have three very different tools: 1. **Gemini Deep Research**: An autonomous multi-step research agent available in Google's consumer products (Gemini, NotebookLM, Google Search, Google Finance) and via the Interactions API for developers. It iteratively plans, searches, reads, and generates a structured, cited report. 2. **Perplexity Deep Research**: A consumer-facing autonomous research mode that performs dozens of searches, reads hundreds of sources, reasons through them, and generates comprehensive reports in 2-4 minutes. Free for all (with daily limits) and unlimited/high-volume for Pro subscribers ($20/mo). 3. **Sacra**: A private market research and data platform (not an autonomous web research tool itself, but *data infrastructure* for AI agents). They write premium equity research, gather financial models, valuation data, and expert interviews on growth/pre-IPO companies. Crucially, they have pivoted to being "agent-ready" by providing an **MCP (Model Context Protocol) server** and APIs. This allows models like Claude and ChatGPT to query their structured private market data instead of relying on "generic webspam". Let's search for **Google Scholar** and how it fits into the AI research landscape, or how other academic/scientific AI research tools like Consensus, Elicit, or SciSpace are positioning themselves. Let's search for "Google Scholar AI features" or "Google Scholar vs Elicit Consensus". Let's do a web search.

Let's fetch `https://aarontay.substack.com/p/google-scholar-vs-other-ai-search-tools` to read the analysis by Aaron Tay, a well-known academic librarian and expert in academic search engines, comparing Google Scholar with AI search tools (Undermind, Elicit, SciSpace, Consensus). This is highly valuable context for understanding features, technology, data sources, and the market map.

Ah, the fetch failed due to a timeout. No problem. Let's do a search to see if we can find other articles or discussions about Aaron Tay's comparison, or let's search for "Google Scholar vs Elicit" or "Undermind vs Google Scholar" to understand how these academic AI research tools differ from Google Scholar. Let's perform a web search.

Let's search for "Undermind.ai" and "Elicit.com" to see how they work. Undermind.ai is mentioned as "the slow but powerful specific searcher". Let's search for Undermind.ai features, pricing, and technology to see how it operates. Let's do a web search.

Ah, the search budget is exhausted, which is perfectly fine because we have a wealth of incredible data points from our searches and fetches! Let's analyze what we have. Let's do a quick breakdown of the market for AI research tools based on our findings: 1. **General/Web Autonomous Research Agents** - **Gemini Deep Research** (Google): Built on the Interactions API and Gemini 3 Pro reasoning core. It iteratively plans, queries, reads, and synthesizes multi-step research tasks. It is optimized to navigate deep into websites and generate comprehensive documents with detailed citations. - *Pricing/Model*: Available via API (based on token consumption of models + tool usage) and consumer apps like Gemini Advanced ($20/mo), NotebookLM, and Google Finance. - *Data Sources*: Public web data accessed via Google Search and user-provided files (PDFs, CSVs, etc.). - **Perplexity Deep Research**: Consumer-facing, autonomous research tool. It conducts dozens of searches, reads hundreds of sources, and writes a comprehensive report in 2-4 minutes. - *Pricing/Model*: Freemium model. Free for all with limited daily queries; unlimited/high volume for Pro subscribers ($20/mo or $200/yr). - *Data Sources*: Live web crawling and content aggregation. 2. **Specialized Academic/Scientific Literature AI Search Engines** - **Google Scholar**: The gold standard legacy search engine based on lexical (keyword) indexing of the academic record. It has massive coverage but lack autonomous synthesis or deep reasoning natively (though it has added minor AI features like PDF reading/summarization). - **Undermind.ai**: Created by MIT researchers. It is a "slow but powerful" literature search tool that autonomously reads hundreds of papers, traversing citation graphs and performing successive searches to find highly precise, relevant insights. - *Pricing/Model*: Freemium. Free version has 5 searches/month (analyzing 50 papers each). Paid/Pro subscription allows deeper analysis (150+ papers per search). Also sells institutional/enterprise licenses (e.g., SMU, universities). - *Data Sources*: The global scientific/academic record (millions of papers, open access and publisher databases). - **Elicit / Consensus / SciSpace**: AI-powered academic search tools designed to answer questions directly from scholarly literature, extract key data points into structured tables, and summarize consensus across papers. - *Pricing/Model*: Subscription-based (SaaS) and enterprise plans. - *Data Sources*: Academic paper corpuses (Semantic Scholar, etc.). 3. **Specialized Private Market/Equity Research Infrastructure** - **Sacra**: A private market research and data provider that has pivoted to being "agent-ready". They write institutional-grade equity reports and compile private company financials (revenue, valuation models, expert interviews). Instead of building a generic agent, they provide an **API** and a **Model Context Protocol (MCP) server** for ChatGPT, Claude, and Cursor. This enables developers and professional investors to bring high-quality, structured private market data directly into their existing LLM workflows. - *Pricing/Model*: Standard ($50/mo for individuals, 500 tasks), Pro ($350/mo for teams, 3500 tasks, MCP/API access), Platform ($1,500/mo, 25,000 tasks, API/MCP), and Enterprise. - *Data Sources*: Proprietary equity research, expert interviews, financial filings, and custom-curated revenue/valuation models of growth and pre-IPO companies. Let's organize these into structured, distinct findings. We want to write one finding per distinct theme or story worth persisting. Let's write: 1. A note on **Gemini Deep Research** (features, technology, API, pricing). 2. A note on **Perplexity Deep Research** (consumer-centric, freemium model, capabilities). 3. A note on **Sacra's agent-ready pivot** (MCP, API, private market data infrastructure). 4. A note on **Academic AI Research Tools** (Google Scholar vs Undermind, Elicit, Consensus). 5. A note on **Market Mapping and Positioning for Hey, Lefty** (how a research assistant fits into this landscape, based on these findings). Let's write the first note. I will formulate a clean title and slug, write a comprehensive markdown body, cite the sources, and use quotes. Let's start with `gemini-deep-research-agent`. No YAML frontmatter. Let's write.

Now let's write a note on **Perplexity Deep Research**, focusing on its consumer-centric, high-speed autonomous research and its freemium business model. Let's write.

Now let's write a note on **Sacra**, focusing on its pivot to "agent-ready" private market data infrastructure via APIs and the Model Context Protocol (MCP) server. This is a crucial, super-smart pivot. Let's write.

Now let's write a note comparing **Google Scholar** with modern **Academic/Scientific AI Research Tools** like Undermind, Elicit, and Consensus, which are revolutionizing how literature reviews and scientific synthesis are conducted. Let's write.

Now let's write a highly strategic note on **Market Mapping and Positioning for Hey, Lefty**, based on these findings. This will help the user see exactly where the market is and where Hey, Lefty should fit. I will use wikilinks to link the other findings. Let's write.

All notes have been successfully written! I will now call `finish_cycle` to complete this research cycle and save these findings.

I have completed the research cycle and mapped the market for autonomous and AI research tools across the general web, premium financial data, and academic literature spaces. Here is an overview of the findings captured during this cycle: ### 1. General Horizontal Web Agents * **Gemini Deep Research (Google)**: An autonomous multi-step reasoning agent powered by a specialized Gemini 3 Pro core. It iteratively plans, queries, reads, and synthesizes reports. It is available to developers via the new **Interactions API** (priced on token + tool consumption) and integrated into consumer/enterprise apps like Gemini Advanced ($20/mo), NotebookLM, and Google Finance. * **Perplexity Deep Research**: A high-speed, consumer-centric autonomous agent that runs sequential searches and reads hundreds of sources to generate structured reports in 2–4 minutes. Offered via a **freemium** model (limited free queries daily, unlimited/high-volume for $20/mo Pro subscribers). ### 2. Premium Private Market Data Infrastructure * **Sacra**: A private market equity research and financial data platform specializing in growth-stage and pre-IPO technology companies. Rather than building a generic browser agent, Sacra has pivoted to being **"agent-ready."** They launched an **MCP (Model Context Protocol) server** and API, enabling ChatGPT, Claude, and local developer agents (like Cursor) to query their premium company profiles, revenue/valuation models, and expert interviews directly—bypassing low-quality web search results. * **Pricing/Model**: High-value SaaS subscriptions spanning **Standard ($50/mo)**, **Pro ($350/mo)**, and **Platform ($1,500/mo)**, which meter access by seats and programmatic "tasks" (API/MCP calls). ### 3. Specialized Academic & Scientific Literature * **Google Scholar**: The legacy baseline index for academic papers, relying on keyword searches and citation counts. * **Undermind.ai**: Created by MIT researchers, this is a "slow but powerful" autonomous literature search tool that reads hundreds of papers in full and traverses citation graphs to automate literature reviews. (Freemium: 5 free searches/month; paid/institutional licenses). * **Elicit / Consensus**: SaaS platforms focused on structured extraction (Elicit maps paper details into comparison tables) and summarizing scientific consensus (Consensus's "consensus meter"). --- ### Strategic Positioning for Hey, Lefty We have compiled these findings into a comprehensive market map and outlined three distinct positioning strategies for **Hey, Lefty**: * **Strategy A (The Agentic Research OS)**: Positioning Hey, Lefty as a multi-source orchestrator. Instead of just searching the web, it connects to and coordinates external MCP servers (like Sacra, your internal database, or academic APIs) to run deep, unified research cycles. * **Strategy B (The Professional Analyst-in-the-Loop)**: Emphasizing the collaborative workspace. Instead of a "black box" report generator, Hey, Lefty acts as an interactive junior analyst that checks in, surfaces structured findings, tracks open questions (threads), and lets the researcher direct the next cycle. * **Strategy C (Verticalized High-Consequence Research)**: Tailoring the platform specifically for corporate strategy, due diligence, and competitive intelligence—emphasizing auditability, verification, and strict anti-hallucination guardrails. The full details, data, pricing structures, and quotes have been saved to your workspace as structured notes: 1. `[[gemini-deep-research-agent]]` 2. `[[perplexity-deep-research-consumer-agent]]` 3. `[[sacra-premium-private-market-mcp-data]]` 4. `[[academic-ai-research-scholar-vs-synthesis]]` 5. `[[market-map-positioning-hey-lefty]]`