The directory says what each tool costs. This page shows the record behind each of those statements: which tool, which exact price claim, which official source it was read on, when it was checked, and how confident we are. You should be able to inspect a claim here instead of trusting a badge. The same records are published as machine-readable data in evidence.json.
This page does not re-argue the methodology — how we verify states the rules. It shows the rules applied, claim by claim.
Every record carries two separate judgments. The status says how far the evidence goes; the method says what kind of evidence it is. Keeping them apart is the whole point: a vendor's own published price can be checked carefully without pretending we tested the purchase ourselves.
| Status | What it means |
|---|---|
| Verified | The evidence supports the precise claim within its stated scope, with no material conditions left out. |
| Qualified | The evidence supports the claim, but important conditions or limitations apply — billing period, per-seat basis, currency, region, or usage limits. |
| Unclear | The available evidence is ambiguous or insufficient to settle the claim. |
| Unverified | No adequate supporting evidence has been established. |
| Outdated | The claim was previously supported but is no longer sufficiently current. |
Method: vendor-stated means the claim is the vendor's own published price, read on its official site on the check date. Method: hands-on tested would mean an editor used the tool and confirmed the claim in practice — no record on this page uses that method yet, and no current listing claims it. A status is always read with its scope: country, currency, billing period, and date. A price checked on November 1, 2026 does not silently stay fresh in January — records are re-checked on the first of each month, and a record past its review date is treated as stale until re-checked.
Pricing claims cover all 21 tools in catalog v1.7.0. Limits, privacy, and feature claims are not yet covered systematically — they are the next tranches.
Tranche 1 covers one claim per tool: the price claim printed in the directory. Every record below is Qualified, vendor-stated: the vendor's official page supports the price, and the conditions that qualify it are printed on the record itself (billing period, per-seat basis, and the tool's limitations). Source links go to the vendor's official site — the same URL the directory links to. The price itself was read on the vendor's pricing page during the re-check dated on the record; the catalog does not store a separate pricing-page URL, so this page does not pretend to link one. If a vendor's page does not disclose a price, we do not infer one from a third-party site — that record would read Unclear or Unverified, with the reason stated.
Price claim: Free $0; Go $8/mo; Plus $20/mo; Business $25/seat monthly or $20/seat billed annually (2-seat minimum); Enterprise custom.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Free and Go may show ads; Plus has usage caps at high demand; a subscription does not include API use, which is a separate metered bill.
Source: OpenAI official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Free $0; Pro $20/mo or $17/mo billed annually; Max from $100/mo; Ultra $200/mo; Enterprise $20/seat/mo plus API usage.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Usage limits reset on a rolling 5-hour window plus weekly caps, shared across web, mobile and code; hitting a limit means waiting, upgrading, or buying credits.
Source: Anthropic official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Free $0; Google AI Plus $4.99/mo; Pro $19.99/mo; Ultra $99.99–$199.99/mo; Enterprise Business from $21/seat/mo.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Free gets only the Flash-Lite model, and usage is stated as multiples of Free rather than fixed quotas; much of Pro/Ultra's price is bundled storage and extras you may not need.
Source: Google official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Copilot Chat included with an eligible Microsoft 365 plan; Business Standard with Copilot $23.50/user/mo, Premium $32.00/user/mo (yearly billing); Copilot Business add-on $18/user/mo (shown from $21).
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Copilot inside Word, Excel, PowerPoint, OneNote and Teams needs a paid license — free Chat does not include it; advanced AI experiences are billed by usage on top of the license.
Source: Microsoft official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Chat app free, no paid tier; API pay-as-you-go per 1M tokens — e.g. Flash output $0.60 off-peak / $1.20 peak, Pro output $1.98 off-peak / $3.96 peak.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: No paid tier buys priority in the chat app — heavy users can hit throttling with no upgrade path; the API generates no images, video or audio.
Source: DeepSeek official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Free $0; Pro $17/mo billed annually ($20 monthly); Max $167/mo billed annually ($200 monthly); Enterprise Pro $34/seat/mo billed annually.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Free is search-only with basic models and limited daily use; agent tasks consume credits and pause when they run out; annual rates are paid up front.
Source: Perplexity official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Basic free; Pro $16.99/mo or $8.33/mo billed annually; Business $19.99/mo billed annually ($24 monthly, first 3 months discounted); Enterprise custom.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Free gives 300 transcription minutes/month with a 30-minute cap per conversation and only 3 lifetime file imports; minutes are a shared pool with no rollover; API and SSO are Enterprise-only.
Source: Otter.ai official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Free $0; Pro $12/member/mo billed annually; Enterprise custom.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: AI prompts capped at 100/mo free and 2,000/member/mo on Pro — only Enterprise is unlimited; Pro allows only 1 style guide and 1 brand tone.
Source: Grammarly official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Free daily generations (count not published); Standard $9.99/mo (2,000 credits); Pro $19.99/mo (4,000); Pro Plus $49.99/mo (10,000); Premium $199.99/mo (50,000).
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Everything premium runs on generative credits that do not roll over, and video burns them fast — check the credit model on the official page before buying for video-heavy work.
Source: Adobe official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Free $0 (100 tasks/mo); Professional from $19.99/mo; Team from $69/mo; Enterprise custom.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Billing is per task — every successful step costs one, and AI steps cost more, so multi-step Zaps burn the allowance fast; with overage off, Zaps pause at the limit.
Source: Zapier official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Free $0; Plus $12/member/mo ($10 billed annually); Business $24/member/mo ($20 billed annually); Enterprise custom.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Free and Plus include only a trial of Notion AI — full AI (Agent, Meeting Notes, Enterprise Search) needs Business or higher; Custom Agents are metered at $10 per 1,000 credits on top.
Source: Notion official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Free $0; Pro $18/mo or $144/yr for one person; Business $25/person/mo or $250/person/yr; Enterprise custom.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Free gives 5 GB storage and up to 20 Standard or Premium AI uses; Pro is a single seat — a team needs Business.
Source: Canva official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Basic $10/mo ($8/mo billed annually); Standard $30/mo ($24); Pro $60/mo ($48); Mega $120/mo ($96). Extra Fast GPU time $4/hour.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: No free tier or trial; Basic has no Relax mode and only 3.3 Fast GPU hours/month; private (Stealth) mode is Pro/Mega only; companies over $1M/year revenue must buy Pro or Mega.
Source: Midjourney official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Pro $69/seat/mo or $59/seat/mo billed annually; Business custom. 7-day Pro trial, no free tier.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Pro includes only 1 seat and caps at 2 Brand Voices, 5 Knowledge assets, 3 Audiences; API access and enterprise controls are Business-only, with pricing undisclosed.
Source: Jasper official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Free $0 (1,000 credits/mo); Core $16/mo monthly or $12/mo billed annually; Pro $28 or $21; Teams $51 or $38; Enterprise custom (prices at the 10,000-credit setting).
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Every module action consumes credits, so one run can cost from 2 to thousands of credits; scenarios stop when credits run out, and unused credits expire at term end.
Source: Make official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: n8n Cloud: Starter €20/mo billed annually (2,500 executions/mo; US pages list $24/mo); Pro €50/mo billed annually (10,000 executions/mo; US pages list $60/mo); Business €667/mo billed annually (self-hosted license); Enterprise custom. Self-hosted Community edition free under the fair-code Sustainable Use License — you pay for your own server. Cloud starts with a 14-day free trial; there is no permanent free Cloud tier.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Cloud billing is per execution — one full workflow run counts once however many steps it has, which is kind to complex workflows, but a workflow that polls every few minutes burns the monthly allowance fast. Self-hosting is real work: Docker/server setup, updates, backups and security are on you, and the free Community edition drops SSO, projects, sharing and Git version control. The Community edition is fair-code (Sustainable Use License), not open source: free for your own internal business use, but you may not host n8n for others for money or resell it without a commercial license. AI steps run on your own model accounts, so model bills come on top.
Source: n8n GmbH official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Free with a Google account; higher limits via Google AI plans — Plus $4.99/mo, Pro $19.99/mo, Ultra from $99.99/mo. No standalone NotebookLM subscription.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Google's page only promises 'expanded access' per tier — exact notebook and source limits per tier are not published there; answers come only from documents you upload.
Source: Google official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Free $0 (generous weekly limit of 100M tokens, as metered on a free account; Meta does not publish the figure); Power $20/mo (500M tokens/week); Maximum $100/mo (3B tokens/week).
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Available only in countries where Muse has launched, and subscriptions are in limited testing; 18 or age of majority; sensitive actions (payments, messages on your behalf) wait for your approval; the free weekly limit can run out on long tasks.
Source: Meta official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Software free (MIT license, self-hosted) — you pay for the model you connect. Optional Nous Portal: Free $0 (free models only, $0 credits); Plus $20/mo ($22 in monthly credits, $10 rollover cap); Super $100/mo ($110 credits, $50 cap); Ultra $200/mo ($220 credits, $100 cap).
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: You host and secure it yourself: setup is an installer and terminal work, the model bill and any server are on you, and an always-on agent with shell access needs sandboxing and command approvals — never expose its dashboard publicly without authentication.
Source: Nous Research official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Free and open source (MIT license); no paid tier and no hosted service — you pay only for the model you connect (API usage or an existing subscription) and for any computer or VPS it runs on.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Self-hosted and technical: installs, updates and security are yours. It can run shell commands and touch your files, and its default setup assumes a trusted operator with sandboxing off — isolate it (separate machine, VPS or Docker) before connecting real accounts.
Source: OpenClaw project official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Price claim: Standard $20/mo (4,000 credits; about $17/mo billed annually); Customizable from $40/mo (8,000 credits; $34/mo billed annually); Extended $200/mo (40,000 credits; $167/mo billed annually); Team from $20/seat/mo. Manus's help docs reference a free plan, but no free plan was shown on the public pricing page when checked.
Status: Qualified — the vendor's official page supports this price within the scope stated above, but conditions apply (see below). Method: vendor-stated, not hands-on tested.
Conditions: Everything runs on credits, and the cost of a task is not shown before it runs; monthly credits do not roll over, and a single deep-research task can burn roughly 900 credits — about a quarter of the entry plan. Work can stop if credits run out mid-task.
Source: Manus AI official site → · Checked: 2026-11-01 · Next review due: 2026-12-01
Later tranches add free-tier limits, data retention and training, commercial-use rights, and agent permissions as separate claim types, each with its own source and status. When a price changes, the old record is marked Outdated, a replacement record is created, and the change is entered in the Change Ledger — history is preserved, not overwritten. Spot an error, a broken source, a changed price, or a regional difference? Corrections are welcome from anyone — send them to [email protected]. No payment is ever required to correct a factual error, and a correction never buys a better listing.
Browse the full directory → How we verify → Find your first tool →