Research-Tool Market Segmentation & Price Points, Mid-2026 — and the Verified Capability Limits

Research-Tool Market Segmentation & Price Points, Mid-2026 — and the Verified Capability Limits

Revised 2026-10-14: adds Paperguide's fully published academic evidence-synthesis ladder (fetched live from paperguide.ai/pricing) — the cleanest public example yet of credit-metered agent work, audit-grade features as the top-tier differentiator, and a destination platform exposing its own MCP backend at the free tier.

The brief asks who the AI research tools sell to, at what price, and on what differentiated capability. Three sources fetched last cycle (2026-09-17) gave a procurement-grade snapshot of the market's segments (consumer answer engines, free incumbents, deep-research subscriptions, vertical audit-grade tools, metered API backends). This cycle extends the price-point record into the academic evidence-synthesis segment with Paperguide, whose self-description is: "Find, organize, and synthesize research from first search to cited draft, with rigorous, auditable systematic reviews when you need them."

Paperguide's published ladder (fetched 2026-10-14):

• Free — 2,000 AI credits/month "across Research Agents, Research Reports, and AI Writer"; Research Agent; limited reference manager; Paperguide MCP; plagiarism checker 4,000 words/mo; 10 Search API requests/mo.
• Plus — $12/seat/mo (billed annually) — 12,500 credits; "Systematic reviews up to 1,000 papers"; unlimited reference manager; admin panel, seat management.
• Pro — $29/seat/mo — 50,000 credits; systematic reviews up to 5,000 papers.
• Evidence — $89/seat/mo — 150,000 credits; "PRISMA grade dual reviewer screening"; dedicated support channel.
• Enterprise — custom — systematic reviews up to 20,000 papers; custom credits and API requests.

Three map-relevant signals in one vendor's rate card:

1. Credit metering as the unit of agent work. Every tier meters "AI credits" across agents/reports/writer — the same pattern as Exa Agent Ultra's per-run budget, Parallel's per-request CPM, and OpenAI's dots: agentic research is sold as a consumable allowance, not a seat-flat feature¹.
2. Audit-grade features carry the top tier. The $89 "Evidence" tier's differentiator is PRISMA-grade dual reviewer screening — methodology compliance (not raw intelligence) is what commands 7.4x the entry seat price.² A customer testimonial quantifies the value prop: "it cut our abstract screening from about three weeks to one, and the decisions came out more consistent, not less." This slots Paperguide beside the audit-grade verticals in The Audit-Grade Vertical: Research Agents That Sell "Show Your Work" — Lumivero, Menos AI, HITZ and the academic layer in The Academic Deep-Search Layer: Undermind's $16/mo Agent vs. Ai2's Free, MCP-First Asta — and the Independent Spot-Checks of Its Recall Claim.
3. Destinations ship MCP backends now. Paperguide MCP appears in the free tier — an academic destination platform exposing its corpus/workflow to external agents³, continuing the destination→backend convergence tracked in Destination vs. Backend: The Research-Tool Market Bifurcates — and Both Layers Converge on MCP and The Licensed-Access MCP Layer: Publishers and Database Owners Sell Agent Entry to Their Corpora — With Published Price Points.

The seat ladder also spans the full academic buyer arc — students ($12) → faculty/small teams ($29) → review teams ($89) → institutions (custom) — with university logos (MIT, Oxford, Berkeley, Duke, NYU) as social proof, the same land-and-expand shape the vertical layer uses elsewhere (cf. OpenEvidence in the EHR: Vertical Research Tools Win by Embedding Into the Institution's System of Record — and Now the Hosts Are Embedding Too).