Head to head · Lead search · October 2026 research run

Leadsets vs Retriever

Leadsets scores 62.9 (B) on agent readiness against Retriever's 32 (F), and leads in 5 of 7 scored categories. Retriever leads on reliability and schema & documentation. Both do lead search.

Best lead, contact and company data APIs for AI agents · All 276 leads comparisons

Which one, for what

Leadsets B

Best for Prospect search, contact enrichment with verified work emails and company-site scraping for an agent on a flat monthly plan rather than credits.

Ahead on

  • Agent ergonomics, 90 against 75
  • Security & auth, 67 against 23
  • Payments & pricing, 50 against 30
  • Maintenance & community, 80 against 5
  • Transparency & trust, 67 against 49

Also in its favour

  • Free to start without a card
  • No incidents deducted, where Retriever loses 15 points for them

Watch for

No status page, SLA or incident history found, and the terms promise no uninterrupted service unless agreed in writing

Retriever F

Best for A sales team that wants an agent to build qualified lead tables and run outreach from Claude, Cursor or Codex.

Ahead on

  • Reliability, 45 against 40
  • Schema & documentation, 82 against 58

Watch for

Its sequences send LinkedIn invitations, messages, visits and likes from connected accounts automatically, which LinkedIn forbids, so customers' accounts can be restricted

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Maintenance & community Leadsets +75

Leadsets 80

The product is about two weeks old. The TypeScript SDK 0.2.1 was tagged on 22 November 2026, the day of this check (30). Four SDK tags (0.1.0 and 0.1.1 on 11 November, 0.2.0 on 19 November, 0.2.1 on 22 November) and four PyPI releases from 11 to 19 November, all within 90 days (20). A support email with response times by plan and an API changelog with no entries yet. GitHub issues weren't read (7 of 25). In the official MCP registry as io.github.leadsets/leadsets since 14 November 2026, under a namespace proved through GitHub, and current official SDKs in two languages (15). The SDK has no runtime dependencies, CI on Node 20, 22 and 24, and npm trusted publishing with provenance, and the CI run results weren't read (8 of 10). 30 + 20 + 7 + 15 + 8 = 80.

Retriever 5

No changelog, release notes or dated API change found, and the docs carry no dates; the blog's posts are marketing articles, not release notes (0 and 0). Support is an email address and a demo booking, with no public changelog or community channel that answers (5 of 25). Not in the official MCP registry and no official SDK (0). No package to check (0). The domain dates from 7 April 2026.

Security & auth Leadsets +44

Leadsets 67

Bearer keys with a full, data or read access level, an expiry of never, 30, 90 or 365 days, hashed storage, revocation within a minute and a documented rotation procedure. MCP clients can use OAuth with PKCE S256 and dynamic client registration, with no scopes in the server metadata. No query-string key documented (30). A read-level key can call only the usage endpoints, and every data tool reads rather than writes, so there is nothing destructive to confirm. Members see and revoke only their own keys (16 of 20). scrape_company and scrape_jobs return text from company websites, and no prompt-injection guidance was found (3 of 15). A request log in the dashboard, an optional X-Leadsets-Client label, rate-limit headers and request_id on every response (12 of 15). /.well-known/security.txt answered 404. The SDK repository's SECURITY.md names security@leadsets.ai with acknowledgement within 3 business days. No bug bounty, and SOC 2 Type I readiness has a target of Q3 2027 (6 of 20). 30 + 16 + 3 + 12 + 6 = 67.

Retriever 23

The MCP server's OAuth 2.1 issues an ordinary API key, and keys have no scopes or expiry and work as their creator in that organisation: plain revocable keys (20), less 10 because the MCP docs offer the key in the URL as ?api_key= (10 of 30). No read-only mode; destructive hints and explicit-request instructions on cancels and deletes, and the outreach toolset can be switched off (8 of 20). The tools return scraped pages, social posts and inbox replies, with no prompt-injection guidance found (0). Credit charges per run and a credits ledger by source, but no audit log documented (5 of 15). No security.txt (404), disclosure policy, bug bounty or certification found (0).

Schema & documentation Retriever +24

Leadsets 58

No OpenAPI file or other machine-readable contract found in the docs or llms.txt, and the MCP server answers initialize with 401 without a key, so its tool schemas couldn't be read (0). llms.txt, llms-full.txt and a Markdown twin of every page (10). The REST reference says what each route does, what it costs and what it returns, with little on when not to use it, and the MCP definitions are unread (12 of 20). People search inputs are typed with enums for seniority and function, ranges for limit (1 to 100) and max_age_days, ISO country codes and booleans, with one nested priorities object (13 of 15). cURL, TypeScript and Python examples and a documented response, plus an errors page that lists 11 types with what to do (15). /v1 in the path and a dated, semver CHANGELOG.md for the TypeScript SDK, but the API changelog had no entries on 22 November 2026 (8 of 15). 0 + 10 + 12 + 13 + 15 + 8 = 58.

Retriever 82

OpenAPI 3.1 file, version 1.0.0, about 51 paths, linked from llms.txt (25). llms.txt and a Markdown copy of every docs page (10). The tool catalogue says when to use most tools and when not to (cancels only on the user's request, slow services only in a batch) (17 of 20). The OpenAPI file has enums and required fields; the MCP input schemas weren't read, since tools/list needs a key (11 of 15). An errors page with stable codes and the action for each, plus examples in the docs (14 of 15). Routes are versioned under /v1 but there's no public changelog (5 of 15).

Payments & pricing Leadsets +20

Leadsets 50

No x402, MPP or L402. An exhausted Free meter answers 402 free_records_exhausted with no payment requirements (0). Plan prices and their per-window limits are public, with no per-record price, and scraping units are priced per pack ($9 for 1,000, $19 for 5,000, $49 for 20,000), so plan-only pricing with a per-unit add-on (20 of 30). Free plan of 1,000 enrichment records, 100 search records and 1,000 data calls, one time, "no card required" (30). A person signs up in a browser and makes the key in the dashboard (0). 0 + 20 + 30 + 0 = 50.

Retriever 30

No x402, MPP or L402 (0). Credits per action and the dollar price per credit ($0.0134 on Starter) are public without a login (30). No free tier or trial without a card found; the terms say only that free tiers may be offered (0). Access needs a browser signup and a subscription, so an agent can't onboard itself (0).

Transparency & trust Leadsets +18

Leadsets 67

Closed service with terms naming Mobylogix LLC under Texas law. The SDK is MIT (15). The privacy policy, terms, DPA and compliance page say where dataset records come from (licensed suppliers and permitted public sources such as company websites), what is held, that uploaded files are deleted 30 days after a job, and how to opt out. The suppliers aren't named, and the privacy policy says analytics run only on the public website while the compliance page lists PostHog and Google Analytics as "Product analytics (customers only)" (20 of 30). Breaking changes get 90 days' notice by email and material changes to plans or terms 30 days, with no dated notices yet (15 of 20). Subprocessors named (Vercel, Supabase, Stripe, Resend, PostHog, Google Analytics), all in the US, with 30 days' notice before adding one (18 of 20). 15 + 20 + 15 + 18 = 68.

Retriever 49

Closed source with clear terms, last updated 30 August 2026 (15). The privacy policy states retention (deletion within 30 days of closing an account) and names some subprocessors, but says nothing about the LinkedIn and email accounts customers connect, the people in lead tables or the data providers named on the pricing page (12 of 30). The terms say the service may be changed or removed "with reasonable notice" and material changes to the terms get 15 days' notice; no deprecation dates found (8 of 20). Subprocessors are given as examples (Vercel, Railway, Supabase, Anthropic, OpenAI) with no data locations, beyond Standard Contractual Clauses for transfers (8 of 20).

Agent ergonomics Leadsets +15

Leadsets 90

Eight MCP tools per the docs, with sizes unread, and REST responses sized by limit and a count_only mode that returns only a count, but no field selection found (20 of 25). Cursor pagination, typed filters, max_age_days, exclude_stale and has_verified_email (20). Every error has a status, a type, a request_id, the meter and charged, and the docs say whether to retry each one (20). Idempotency-Key on every paid call for 24 hours, with 409 idempotency_conflict for a reused key with a new body. MCP tool annotations couldn't be read (15 of 20). One required header and defaults such as limit 25, and official TypeScript and Python SDKs that retry 429 and 5xx (15). 20 + 20 + 20 + 15 + 15 = 90.

Retriever 75

67 tools in the catalogue (5), with 5 back because the outreach tools are a toolset shown only when the organisation has a sender and switchable in settings (10 of 25). Limit, offset and cursor pagination, up to 1,000 rows a call, and the first 20 rows returned with a finished run (18 of 20). Stable error codes with the action to take (20). Idempotency-Key on run creation, and readOnlyHint, destructiveHint and idempotentHint on the tools (20). Few required parameters (a run needs only a prompt), but no official SDK, only a Node client in a recipe (7 of 15).

Score by category

CategoryWeightLeadsetsRetrieverEdge
Reliability16%4045Retriever +5
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%5882Retriever +24
Agent ergonomics13%9075Leadsets +15
Security & auth14%6723Leadsets +44
Payments & pricing10%5030Leadsets +20
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%805Leadsets +75
Transparency & trust7%6749Leadsets +18
Negative events≤150-15
TotalB 62.9/100F 32/100

Facts side by side

FactLeadsetsRetriever
KindHTTP APIHTTP API
VendorLeadsetsRetriever
Hosted endpointhttps://mcp.leadsets.aihttps://api.retriever.run/mcp
TransportsStreamable HTTPStreamable HTTP
AuthOAuth or keyOAuth or key
PricingFreemiumPaid
x402nono
Tools exposed867
Read-only variant documentednoyes
llms.txtyesyes
MCP registryio.github.leadsets/leadsetsnot listed
Last release2026-11-22none
Terms last updated2026-11-152026-08-30
Privacy policy last updated2026-11-192026-08-30
Customer content may train modelsnot found in the textnot found in the text
Terms restrict automated accessnot found in the textyes
Terms restrict benchmarkingnot found in the textnot found in the text
Terms or service can change without noticenot found in the textnot found in the text
Arbitration or class-action waivernot found in the textnot found in the text

Verdicts

Leadsets

Agents get eight MCP tools or REST routes with typed errors that say whether to retry, idempotency keys on paid calls, read-only keys and a free tier with no card. The product is about two weeks old, with no status page, no OpenAPI file, an empty API changelog and no named data suppliers.

Retriever

Retriever automates outreach from customers' LinkedIn accounts, which LinkedIn's rules forbid. For agents it has an OAuth MCP server, an OpenAPI file, idempotency keys, stable error codes and published credit prices, but no status page, changelog, security contact, free tier or named company behind it.

Before you call either

Leadsets

  1. Call get_usage before a large run. It costs nothing and returns every meter with its reset time
  2. Send an Idempotency-Key on every paid call, and reuse it on a retry so records are counted once
  3. If a 429 has type weekly_limit, stop until resets_at. For window_full and too_many_concurrent, wait for Retry-After
  4. Use a read-level key for monitoring only. Data tools answer 403 forbidden to it
  5. Treat text from scrape_company and scrape_jobs as untrusted website content, since no injection guidance was found

Retriever

  1. Don't enrol leads in a sequence with LinkedIn steps without the user knowing LinkedIn forbids automated activity and can restrict their account
  2. Send an Idempotency-Key with each run and replay with the same key after a timeout, so a retry doesn't start a second billed run
  3. Run getEmail, getPhoneNumber and webResearch through service_batch_execute, even for one person; service_execute refuses them
  4. Check credits before a paid run, and call service_inspect for a service's thingsToKnow and credit cost before executing it
  5. On 429 wait for Retry-After; at most 5 runs execute per organisation, so a queued run starts on its own and shouldn't be created again

Questions

Which is better for AI agents, Leadsets or Retriever?

Leadsets scores 62.9 (B) on agent readiness against Retriever's 32 (F), and leads in 5 of 7 scored categories. Retriever leads on reliability and schema & documentation.

Do Leadsets and Retriever need an API key?

Both take an API key or an OAuth sign-in.

Can an agent call Leadsets and Retriever without installing anything?

Yes. Leadsets has a hosted endpoint at https://mcp.leadsets.ai and Retriever at https://api.retriever.run/mcp.

Other comparisons with Leadsets or Retriever

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.