The provider for
YouTube sponsorships.
Interface optional.
Research, vet, and buy YouTube sponsorships from your terminal. Against 1.3 million tracked channels, 61,008 brands, and TL's opted-in inventory: 11,537 channels in the Media Selling Network and 170 elite channels in the ThoughtLeaders Partner Program. Install in 30 seconds. Query at agent speed. Pay only for results.
$ tl brands Surfshark --json | jq '.results[0]' ● Pulling brand intelligence... (5 credits/result) ● Matching against 1,299,244 channels and 61,008 brand profiles... ✓ 58 sponsorships · 15 channels · 43 repeat deals · $775 avg · 1.8s brand: Surfshark total_sponsorships: 58 (last 12 months) unique_channels: 15 repeat_rate: 74% ← strongest retention signal in the VPN category avg_deal_price: $775 top_channels: Hustl, Black Pigeon Speaks, Behemeth...
Agent accessibility is the new API.
APIs promised interoperability for 20 years. 99.9% of users never touched one because they required a programmer. Agents changed that. A CLI your agent can call is an API your non-technical user can finally use.
Speed is the feature.
Agents run at 200+ tokens/sec against a text CLI. Against a browser UI, they screenshot, reason, click, wait, screenshot again. Pick the interface that keeps up with the model.
The UI is optional. The data is the product.
Our moat isn't the dashboard. It's 7 years of matched sponsorships, 1.3M tracked channels, and outcomes data nobody else has. The CLI exposes that moat directly. Bring your own interface.
Works with every agent harness and every shell.
One binary. Every harness. No vendor lock-in.
Same question. Same creator. 2h 14m → 1.8s.
Before
- — 47 browser tabs open
- — A Google Sheet with 412 rows of scraped channels
- — An 80-message Slack thread with your AM
- — A Notion doc titled "Q2 sponsorship ideas — FINAL v4"
- — Stopwatch: 2h 14m
- — A sticky note: "is $15k fair? ask eli tomorrow"
Two hours to vet one creator. No idea if the price is fair. No data on repeat sponsorships. No visibility into the evergreen tail. You are guessing.
$ tl channels 12465 --md && tl deals channel:12465 --limit 10 --md ScreenCrush · 2.1M subs · media entertainment ────────────────────────────────────────────── 58 tracked sponsorships · 44 repeat deals (76%) median deal: $4,161 · last 10: Fum 2026-03-25 $4,600 Snakzy 2026-02-05 $2,500 (×4 repeat) Send A Cake 2026-01-27 $3,666 (×3 repeat) Fum 2026-01-27 $4,596 (×3 repeat) ... credits spent: 27 · time: 1.8s
A 76% repeat rate. The median deal price in black and white. Four repeat brands jumping off the page. You know whether $15k is fair in under 30 seconds.
13 real commands. All shipped today.
tl deals 2c / 3c Your sponsorship history back: every deal, cost, price, send date, channel, status. Filter with status:sold brand:"Nike" since:2026-01.
tl channels 3c / 5c Search 1.3M YouTube channels by category, min-subs, language. Real profile data, not YouTube API echoes.
tl brands 5c / 8c Brand intelligence on 61,008 brands. Every channel sponsored, every repeat deal, every organic mention — from the Brand Extractor.
tl uploads 1c / 2c Video-level data across every tracked channel: views, likes, comments, brand mentions, evergreenness score.
tl snapshots video 1c/pt Firebolt time-series for any video: view count, likes, comments by age. The raw evergreen data.
tl snapshots channel 1c/pt Historical channel growth: subscribers and views by date, going back to 2021.
tl deals create FREE Create a matched deal proposal directly from the CLI. (And yes — this actually places a real sponsorship. More on that below ↓)
tl reports run varies Re-run any saved AI-built report with new dates. Powered by the same 3-agent review loop TL uses internally.
tl describe FREE Self-serve schema. Every resource, filter, field, and credit cost — the CLI documents itself.
tl ask +2c Natural-language fallback for when you don't have an agent in the loop. Surcharge waived with --llm-key.
tl balance FREE Credit balance and recent usage. Always free.
tl auth / doctor / setup claude FREE PKCE OAuth login, health check, Claude Code plugin installer. All free.
Every command supports --json, --csv, --md, --quiet, --limit, --offset. TTY-detecting: Rich tables in the terminal, raw JSON when piped.
YouTube sponsorships are the only ads that keep paying you after you stop paying.
A Meta ad stops the day your budget stops. A programmatic banner is a one-shot impression. A YouTube integration runs once — and then keeps running. It stays on the channel's page, in their uploads list, in search results, in recommendations. People still find it, still click it, still watch it.
The ad you paid $10k for in January is still collecting views in June, September, and deep into next year. You paid once. You get returns forever.
The problem nobody solved: measuring that tail honestly. You need historical view snapshots, time-series data, and brand mention detection at scale, for years. So we built it.
evergreenness = (views_180d − views_30d)
────────────────────
views_30d
# longform avg: 2.29
# shorts avg: 0.68
# general knowledge longform: up to 12.3
Stored at the article level. Calculated daily over Firebolt's 7.4B-row article_metrics table. Queryable via tl uploads today.
"Every creator tool can tell you view counts. None of them calculate evergreenness at the video level across millions of videos. We do. And now you can query it."
Every query runs against data nobody else has.
The Brand Extractor
We process every brand mention in every video we track — transcripts, descriptions, on-screen — and classify each one: sponsored or organic. That's the engine behind the CLI. Nobody else has it.
$ tl brands --category vpn --window 12m brand channels deals repeat avg ────────── ──────── ───── ────── ───── Surfshark 15 58 43 ← $775 NordVPN 9 9 0 ← $831 ProtonVPN 1 5 4 $3,000 CyberGhost 2 4 2 $1,325 ↳ Surfshark is running the strongest retention program in the category. NordVPN: 9 one-shots, zero repeats.
61,008 brand profiles
Every channel a brand has ever sponsored. Every organic mention. Every repeat. Every dollar figure we've observed across 7+ years. You query it like a database — because it is one.
$ tl brands Manscaped --md total_sponsorships: 55 first_deal: 2020-10-28 most_recent: 2025-08-19 unique_channels: 30 repeat_rate: 45.5% ← strong avg_deal_price: $1,583 recent_deals: • ScreenCrush $2,800 • Capt. Panda $600 • Forrest Kritzer $450 • Theme Park Express $340
The way YouTube sponsorships are bought in 2026.
Every one of these is real. We picked them from actual TL customer conversations.
The Scraped-List Shotgun
You bought a scraped CSV of 2,000 'top fitness channels.' You're cold-emailing 200 of them hoping 5 reply. Half are brand accounts. A third are dead. One is a 2012 talk show clip.
tl channels category:fitness min-subs:100k The Spreadsheet of Truth
A 412-row Google Sheet your AM maintains by hand. Out of date by the time you open it. Three people have write access and none agree on what 'sold' means.
tl deals status:sold since:2026-01 --csv Vibes-Based CPM
The creator's manager says $15k. You have no idea if that's fair. You ask in Slack. Three people give three different answers. You guess and hope.
tl channels 12465 && tl deals channel:12465 --md The Week-One Lie
Your attribution tool reports 'performance' based on the first 7 days. No one checks again. The 85% of views that show up in months 2–12 are never measured.
tl snapshots video <id> --channel <id> Competitor Blindness
Your biggest competitor just sponsored 12 channels in your exact niche. You find out 4 months later when one posts on LinkedIn.
tl brands <competitor> --limit 50 --md The Repeat Blind Spot
You booked $42k on three channels last quarter. Two delivered great results. You have no idea which two. Next quarter you book all three again anyway.
tl deals status:sold --fields channel,price,repeat Every one of these costs you money. Every one is solved in under 30 seconds. That's the entire value proposition.
Most influencer tools are glorified databases.
This one actually books the deals.
You don't just search with the CLI — you buy with it. tl deals create --channel <id> --brand <name> --price <usd> places a real matched deal into the ThoughtLeaders pipeline. Zero credits. RLS-enforced guardrails on the backend. The same pipeline TL's own sales team works out of every day. Under the hood: two tiers of opted-in inventory we've spent seven years building.
Creators who have raised their hand and said "yes, I'll take a sponsorship through ThoughtLeaders." Active inventory across every major YouTube category — gaming, finance, tech, lifestyle, education, entertainment.
6,173 channels joined MSN in the last 90 days alone — the network is more than doubling this year.
$ tl channels --msn category:gaming min-subs:100k $ tl deals create --channel 12465 --brand Acme --price 4600 ✓ Deal placed — free (0 credits) → Track at thoughtleaders.io/adlinks/...BROWSE MSN INVENTORY →
Hand-picked from thousands of monthly collaborations. These aren't channels with big numbers — they've passed a five-stage vetting funnel:
"Your brand's reputation is in safe hands."
$ tl channels --tpp category:tech min-subs:500k $ tl deals create --channel 419872 --brand Acme --price 18000 ✓ TPP deal placed — white-glove handoff → Named AM within 4 hoursLEARN ABOUT TPP →
Both tiers. One CLI. Free to install, free to create a deal, credits only when you read data to decide. No other tool lets you do this — because no other tool has the inventory.
Built for marketing managers who buy YouTube sponsorships.
You run influencer campaigns on YouTube. Not Meta ads. Not programmatic. The real kind — flat fees, script approvals, send dates, creators in your inbox. You spend six figures a year and have no good way to check your work.
This is for you. You shouldn't need to learn a CLI, learn SQL, or hire a data team to answer the obvious questions about the channels you're buying from.
- Creators — Price yourself. See which brands spend in your vertical.
- Talent agents — Negotiate with real data.
- Agencies — Plug it into client workflows.
- Researchers — HBS/MIT academic tier.
- Data partners — Read-only programmatic access to TL's data.
Harvard Business School
Two researchers at HBS and MIT Sloan are building a case study on influencer marketing dynamics, reputational trust, and brand-creator fit. TL's 7 years of matched sponsorship data is the backbone. We shipped password auth specifically so they could sign in without Google SSO.
A developer-education platform
Our first external production customer is a developer-education company that uses tl channels to find authentic creators for their growth team. Read-only DB access through a scoped key, with the same matching models TL uses internally. We're not naming names yet — but they're shipping in production today.
OpenClaw
TL's own AI operator runs this CLI inside TL's internal Slack to run the business every day. Sales team asks OpenClaw to pull deals, vet proposals, draft matches. It's not a demo — it's how we sell.
"A 1-person company with the CLI and a good agent can operate like a 10-person sponsorship shop."
Peter Steinberger shipped 6,600 commits in a single month solo using AI agents. The skill isn't executing tasks anymore — it's designing systems that AI can build, test, and improve. The TL CLI is the sponsorship-ops version of that unlock. Bring an agent. Query the data. Propose the deal. Ship the outreach. Track the evergreen tail. One person. Ten-person output.
Try a query →Install is free. You only pay for queries.
No subscription. No seats. No minimum. Deposit funds, spend credits per result returned, stop paying the moment you stop querying. Deal creation is always free.
| Resource | List | Detail |
|---|---|---|
| tl deals | 2c | 3c |
| tl uploads | 1c | 2c |
| tl channels | 3c | 5c |
| tl brands | 5c | 8c |
| tl snapshots | 1c/point | — |
| tl deals create · proposals | FREE | — |
| tl comments | FREE | — |
| tl describe / auth / balance / doctor | FREE | — |
The CLI is live. Your agent is waiting.
30 seconds to install. Zero credits to prove it works. No subscriptions, ever.