TubeLab — niche research and outlier hunting, and the revenue estimates that are guesses

• TubeLab

  This page

  PricingRival leads

  • Free to start, with tiered paid plans that unlock the data that actually informs a decision
  • The filters that matter most — revenue, RPM and monetisation status — sit above the cheapest tier, so the free plan is a trial of the interface rather than the answer
  • Plans are structured around credits, which fund AI generations and API calls, so heavy programmatic use is metered
  • API access and scriptwriter generations are among the things a higher tier unlocks rather than being available throughout

• vidIQ

  PricingRival leads

  • A usable free tier, with paid plans gated mainly by how much tracking and how many suggestions you get
  • Pricing scales with the plan rather than with traffic, so cost is predictable
  • The free tier is genuinely functional for a single channel, which is why so many creators start here
  • The browser extension is part of what you are paying for, so the value is tied to using the platform interface

• TubeBuddy

  PricingRival leads

  • A free tier, with paid plans historically connected to channel size rather than to usage
  • Bulk tools and testing features are concentrated in the paid tiers
  • Cost is flat per channel rather than metered, which suits an established channel with regular output
  • Most of the workflow value assumes a channel large enough for bulk operations to save real time

• NexLev

  PricingRival leads

  • Subscription-based, oriented around research rather than around optimising an existing channel
  • No meaningful free tier, so it is a considered purchase rather than something to try alongside
  • Bundles its research, automation and programmatic access into the subscription rather than metering them separately
  • Priced for someone treating research as an ongoing part of the workflow

1. TubeLab — official site
2. TubeLab — niche finder
3. TubeLab — outliers finder
4. TubeLab — pricing
5. TubeLab — API documentation
6. vidIQ — official site
7. TubeBuddy — official site
8. NexLev — official site

Free tier

Yes

Pricing summary

The structure is worth understanding before comparing tiers, because the free plan is a demonstration of the interface rather than a usable research setup. What is gated is precisely the data that answers the question you came with: revenue estimation, RPM, monetisation status and audience characteristics. On the free tier you can see that a filterable niche finder exists; you cannot use the filters that tell you whether a niche is worth entering. Above that, paid tiers unlock the full filter sets across both the niche and outlier views, scriptwriter generations, keyword tracking capacity and, at the higher end, API access. The mechanism to watch is credits. Rather than being a flat subscription, credits fund the generative and programmatic work — AI script and idea generation, and API calls — with each tier including a monthly allowance. That means the cost of heavy use scales with what you do rather than being fixed, which is fine if your usage is steady and worth planning for if you intend to drive a lot of automation through the API. Scriptwriter features are among the things reserved for higher tiers rather than available throughout, and API access is not included on the entry paid plan. Tier names, credit allowances and the exact lines between plans change, so the honest advice is to treat the published pricing page as the only current source and this page as an explanation of how the model works. For a creator deciding whether to pay at all, the useful test is whether you would act differently knowing a niche is monetisable — if the answer is yes, the free tier will not tell you and the paid tier probably will.

Reviewed on 19 September 2026 · TubeLab — pricing

• A niche finder that estimates whether a topic can be monetised

  It scans a very large pool of channels and lets you filter them by estimated revenue, RPM and monetisation status, so the question changes from "is this topic popular" to "is this topic worth entering". That second question is the one that matters and the one a view-count leaderboard cannot answer. The estimates are modelled from public data, so treat them as ranking signals for prioritising research rather than as facts about what any channel earns.

• An outlier library filtered by performance against a channel’s own baseline

  This is the most valuable idea in the product. Instead of surfacing the biggest videos, it surfaces videos that substantially outperformed the channel they were published on — which is a much better proxy for a repeatable format, because it controls for the channel already being large. A library of millions of such videos, filtered further by topic, format and whether the content suits a faceless channel, makes it possible to work backwards from what worked to why it worked.

• Title and thumbnail pattern discovery

  Because outliers are searchable by title and by similar videos, the packaging patterns behind them become visible rather than being something you infer from a handful of examples. For anyone who finds packaging harder than production, this turns an intuition into something closer to a checklist — which titles travel in a topic, which thumbnail approaches recur, and what the shape of a format actually is.

• AI idea and script generation grounded in the outlier data

  Rather than a generic prompt, the generation step draws on the patterns found in the outlier library, so the output is tied to formats that have demonstrably worked in that topic. Generation is funded by credits and reserved for higher tiers. It is a drafting shortcut rather than a replacement for a point of view — the useful part is the structure and the angles, not the prose, and the product’s own positioning is that it writes to your voice as a starting point for recording.

• Keyword rank tracking

  Tracked keywords with periodic rank checks let you watch where your videos sit over time rather than only at publication. Rank tracking is one of the few measurable, non-speculative signals in the whole product — a position is a position — which makes it a useful counterweight to the estimation-heavy parts of the toolset. Capacity scales with tier, so the allowance is worth checking against how many videos you actually maintain.

• API and MCP access to the same data

  The research layer is available programmatically, including through an MCP server that lets an AI assistant query real channels, outliers and transcripts instead of guessing from training data. This is the capability most tools in the category simply do not have, and it changes what the product is for: not a dashboard you visit, but a data source a script, an agent or an automation workflow can pull from. API access sits on the higher paid tiers and is metered by credits.

• Filters that describe the content, not just its size

  Beyond revenue, the filters cover whether a channel is faceless or on camera, whether the content appears to be generated, whether it is aimed at children, whether it is long-form or short, and a set of authenticity and consistency signals. For someone deliberately building a channel that does not depend on appearing on screen, the faceless filter is the difference between a research session and a wasted afternoon.

1. 1

   Read every revenue figure as an estimate, because that is what it is

   The monetisation numbers are modelled from public signals, not reported by the channels. That makes them useful for ranking one niche against another, which is what they are for, and useless as a statement of what anyone earns. The practical habit is to use them only for relative comparisons — this niche over that one — and never to put a number in a plan, a budget or a promise.

2. 2

   Prefer performance-against-baseline over raw popularity

   A video with ten million views on a channel with nine million subscribers says something about the channel. A video with a hundred thousand views on a channel with two thousand subscribers says something about the video, and the format, and the idea — which is what you can actually reuse. Whenever you are choosing what to study, filter for outperformance rather than for size, because the second tells you what already won and the first tells you what could.

3. 3

   Treat testimonials and income claims as advertising

   The earnings quoted on sites in this category are individual results from creators who are frequently advocates, affiliates or paid partners, and the tools themselves cannot guarantee outcomes for anyone. That is not a criticism of the product; it is a description of the category. Take the method and leave the numbers, and be especially sceptical of any source that presents an outlier outcome as though it were typical.

4. 4

   Check whether you need channel access at all before granting it

   Research tools work from public data and do not need authorisation to be useful, whereas optimisation tools need it because they act on your videos. If you are only choosing what to make, you do not have to hand over account access to get the answer. Deciding this deliberately rather than by default keeps a durable grant out of the hands of a vendor you may stop paying next quarter.

5. 5

   Test the free tier against the decision you actually face

   Because the useful filters sit above the cheapest plan, running a free trial without a real question will teach you nothing except that the interface works. Go in with a specific choice to make — a niche, a format, a next series — and see whether the free tier helps you make it. If it does not, you have learned that the paid tier is where the value is, which is honest information either way.

6. 6

   Validate with your own publishing before scaling a bet on it

   The single biggest failure mode in this workflow is treating a promising niche as a certainty and committing months to it. The tool narrows the search space; it does not confirm the answer. Publish something small, watch the real retention and click behaviour, and let your own data — which is the only data about your actual audience — decide whether to keep going.

• Creators choosing a niche rather than optimising one

  The clearest fit. If the question in front of you is what to build a channel around, or whether a format is already crowded, the niche and outlier views answer it directly. This is the moment the product is designed for, and most of its distinguishing features are aimed at it rather than at ongoing optimisation.

• Channels that do not want to appear on camera

  The filters distinguish faceless from on-camera content and surface whether a topic works in that format, which is the specific research problem a faceless channel has. Finding out whether anyone has built an audience in a topic without showing their face is otherwise a slow, manual process of opening channels one at a time.

• People who plan by looking at what already works

  If your creative process starts with the market rather than with a personal idea, this suits how you think. The outlier emphasis in particular rewards a habit of working backwards from what travelled to what it was about, and anyone already doing that in spreadsheets and browser tabs will find it much faster here.

• Developers and technical creators who want the data in code

  The API and MCP access mean the research can be automated, queried from an AI agent, or fed into a workflow rather than read off a dashboard. For anyone who would rather script a recurring question than log in and click filters, that is the feature that decides the choice — most of the category offers no comparable programmatic surface at all.

• Small teams where one person owns growth

  Bundling niche research, outlier discovery, ideation and rank tracking behind one subscription is a real saving for a small team that would otherwise buy a research tool and an optimiser separately. It also means one person can cover the whole planning loop without maintaining several dashboards.

• Anyone evaluating a niche before spending months on it

  The cheapest use of a tool like this is to avoid a bad bet rather than to find a good one. Running a candidate topic through revenue estimation and outlier analysis takes an afternoon and can save a quarter, which makes the subscription rational even for someone who ends up publishing only occasionally.

Revenue and RPM figures are estimates, not reported earnings

Monetisation numbers are modelled from public data rather than supplied by the channels, and the product itself describes them as estimates. They are useful for ranking niches against each other, which is their purpose, and unreliable as absolute figures. Never plan around one, and be wary of any review that presents them as earnings.

Outlier analysis describes the past

The library shows videos that outperformed their own channel’s baseline, which is a strong signal about formats and packaging that have worked. It is not a prediction about the next video, and a saturated format still saturates. The tool narrows what is worth trying; it does not tell you what will succeed.

Filters, credits and tier boundaries move

The most consequential filters — revenue, RPM and monetisation status — are above the entry tier, and scriptwriter generations and API access are also tier-dependent. Credits meter the generative and programmatic work on a monthly allowance. These lines change, so check the published pricing page for the current position rather than relying on a description of the model, including this one.

Channel authorisation is not required for the core research

Because the research is built on public platform data, the niche and outlier tools work without granting access to your channel or your analytics. That is a meaningful difference from optimisation-first tools, where authorisation is essential because the features act on your own videos. Connecting anything here is optional, which keeps your performance data out of a vendor you may not keep.

Programmatic access is the structural differentiator

An API and an MCP server expose the same underlying data to scripts, automations and AI agents, with n8n templates documented for common workflows. Most tools in this category are browser-only, so a recurring question has to be answered by hand every time. Here it can be scheduled or delegated, which changes the economics of using the data regularly at all.

What it does not do

It does not optimise videos you have already published, manage metadata across a back catalogue, or replace the analytics available from the platform itself. It does not test thumbnails or automate publishing. It does not predict outcomes, and it cannot promise revenue. And the income figures used in its marketing are individual creators’ claims rather than expectations you should adopt.