Head to head · Lead search · October 2026 research run
Fuse MCP vs Retriever
Fuse MCP scores 46.2 (D) on agent readiness against Retriever's 32 (F), and leads in 4 of 7 scored categories. Retriever leads on reliability, schema & documentation and agent ergonomics. Both do lead search.
Best lead, contact and company data APIs for AI agents · All 253 leads comparisons
Which one, for what
Fuse MCP D
Best for Suited to a salesperson or founder who wants an assistant to build prospect lists, enrich them and run email and LinkedIn outreach inside one Fuse account.
Ahead on
- Security & auth, 43 against 23
- Payments & pricing, 45 against 30
- Maintenance & community, 35 against 5
Also in its favour
- No incidents deducted, where Retriever loses 15 points for them
Watch for
The only OAuth scope is mcp:tools, so a connection can do everything its user can, including sending email and LinkedIn messages
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 50
- Agent ergonomics, 75 against 57
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.
Schema & documentation Retriever +32
Fuse MCP 50
Readings are per the 8 October check, not re-read on 4 November 2026, since FuseAI's terms bar automated extraction from its website and docs. Contract line 0 (was 22 of 25). The MCP tool schemas were never read, because tools/list needs a signed-in account, and the docs only say every tool ships one. The 8 October reading leaned on the REST API's OpenAPI file, a separate surface, and no reading records a docs page that shows or links it, so it is struck for this line, as jambonz's and stable-audio's were. llms.txt, Markdown twins of every page and a skill file (10). The tool reference shows descriptions that state purpose, the preferred tool for a job and when a destructive tool may be called (16 of 20). Enum lookup tools such as get_company_search_enums exist, but input constraints were not read first-hand (9 of 15). Error results are documented with codes and two JSON examples, plus four plain-text failures, and the docs say the codes do not always match the REST catalogue (12 of 15). No changelog was found and the docs say the server does not announce changes to the tool list. The tool reference is dated 7 October 2026 (3 of 15). 0 + 10 + 16 + 9 + 12 + 3 = 50.
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).
Maintenance & community Fuse MCP +30
Fuse MCP 35
Readings are per the 8 October check, not re-read on 4 November 2026. The tool reference is dated 7 October 2026 and records the October 2026 rebuild of the tool list, within 30 days today (30); it drops to 20 after 6 November unless a newer dated change appears. No changelog or release notes found, so three dated entries in 90 days could not be shown (0). No public changelog or issue tracker. Support is by email at support@fuseai.com, not tested (5 of 15). No readable registry page names the server and no official SDK was found (0). No package to assess (0 of 10). 30 + 0 + 5 + 0 + 0 = 35.
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 Fuse MCP +20
Fuse MCP 43
Readings are per the 8 October check, not re-read on 4 November 2026. OAuth authorisation code with PKCE and dynamic client registration, one-hour access tokens and rotating refresh tokens. The only scope is mcp:tools, so every connection has the user's full permissions (22 of 30). There is no read-only mode. Destructive tool descriptions tell the model to get the user's confirmation, but the server enforces none of this and changes apply at once. Billing, API key and team-removal tools are left out (8 of 20). Tools return inbox email, LinkedIn messages and scraped web pages, and the only guidance found is to connect trusted clients and keep approval prompts on (4 of 15). Audit logging is listed for Enterprise and a credit usage ledger exists, with no per-call log for MCP documented (4 of 15). Programme 7 to 5. The home page lists SOC 2 Type I and CASA Tier 2 with no report, trust centre, security.txt, disclosure policy or bug bounty found, and a bare certification claim with no report or reporting route scores 5, as for zamp. 22 + 8 + 4 + 4 + 5 = 43.
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).
Agent ergonomics Retriever +18
Fuse MCP 57
Readings are per the 8 October check, not re-read on 4 November 2026. 290 tools and about half a megabyte of definitions, with no toolsets, dynamic loading or read-only subset. The docs warn that some clients pass only the first 40 tools (5 of 25). List reads return a data array with pagination, searches return a scroll token or cursor, and tools take limits and filters (18 of 20). Rejected calls set isError and return codes such as TOOL_ARGUMENT_VALIDATION_FAILED with the missing field named. Other failures are short plain text (15 of 20). The docs say every tool carries readOnlyHint or destructiveHint, which we could not read from the server. No idempotency keys, and sends are not marked destructive (14 of 20). No official SDK found, and most changes need an id looked up first (5 of 15). 5 + 18 + 15 + 14 + 5 = 57.
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).
Payments & pricing Fuse MCP +15
Fuse MCP 45
Readings are per the 8 October check, not re-read on 4 November 2026. No x402, MPP or L402 (0). Plan prices and the credit cost of each action are public, and get_credit_prices returns them. Top-up prices are shown only in the app and the Free allowance is stated two ways (15 of 20). A Free plan exists at $0. The Solo plan card is where MCP is listed, and we could not confirm that Free includes it or that signup needs no card (15 of 20). A person signs in through a browser to approve each connection (0). 0 + 15 + 15 + 0 = 30. Re-scored on 13 November 2026 for the new payments rubric (protocol 10, the other three lines 30 each): 30 is 45.
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).
Reliability Retriever +5
Fuse MCP 40
Graded as a hosted service. FuseAI's terms bar automated systems or scripts from extracting content from the Services, which they define to include the website, so every fuseai.com page keeps its 8 October reading, not re-read on 4 November 2026. No status page is linked from the site, the docs or the sitemap (0). No readable incident history (5). MCP limits are published at 60 requests a minute and 5,000 a day per user (15). The server answers 429 with RATE_LIMIT_EXCEEDED and the docs say to back off and retry. No Retry-After is documented for the MCP server and there are no idempotency keys, so backoff guidance alone (10). No SLA found (0). The server carries no beta or preview label (10). 0 + 5 + 15 + 10 + 0 + 10 = 40.
Retriever 45
No status page found, linked from neither the site nor the docs (0), so no incident history to read (5). Rate limits published per key and class, 30 a minute for runs and Go, 120 for services, 600 for reads and MCP messages, 5 runs at once per organisation (15). 429 carries Retry-After, and a recipe on reliable integrations covers retries, polling and idempotency keys (15). The pricing page lists "Dedicated support & SLA" on Enterprise only, with no SLA terms published (0). The API and MCP server aren't labelled beta (10).
Score by category
| Category | Weight | Fuse MCP | Retriever | Edge |
|---|---|---|---|---|
| Reliability | 16% | 40 | 45 | Retriever +5 |
| Performance | 10%pending | pending | pending | not scored in this run |
| Schema & documentation | 13% | 50 | 82 | Retriever +32 |
| Agent ergonomics | 13% | 57 | 75 | Retriever +18 |
| Security & auth | 14% | 43 | 23 | Fuse MCP +20 |
| Payments & pricing | 10% | 45 | 30 | Fuse MCP +15 |
| Task success | 10%pending | pending | pending | not scored in this run |
| Maintenance & community | 7% | 35 | 5 | Fuse MCP +30 |
| Transparency & trust | 7% | 52 | 49 | Fuse MCP +3 |
| Negative events | ≤15 | 0 | -15 | |
| Total | D 46.2/100 | F 32/100 |
Facts side by side
| Fact | Fuse MCP | Retriever |
|---|---|---|
| Kind | MCP server | HTTP API |
| Vendor | FuseAI, LLC | Retriever |
| Hosted endpoint | https://mcp.fuseai.com | https://api.retriever.run/mcp |
| Transports | Streamable HTTP | Streamable HTTP |
| Auth | OAuth | OAuth or key |
| Pricing | Freemium | Paid |
| x402 | no | no |
| Licence | Proprietary hosted service under FuseAI's terms of service. No public source repository found | none |
| Tools exposed | 290 | 67 |
| Read-only variant documented | yes | yes |
| llms.txt | yes | yes |
| Last release | 2026-10-07 | none |
| Terms last updated | 2026-07-01 | 2026-08-30 |
| Privacy policy last updated | 2026-07-01 | 2026-08-30 |
| Customer content may train models | not found in the text | not found in the text |
| Terms restrict automated access | yes | yes |
| Terms restrict benchmarking | not found in the text | not found in the text |
| Terms or service can change without notice | not found in the text | not found in the text |
| Arbitration or class-action waiver | yes | not found in the text |
How the scores have moved
Verdicts
Fuse MCP
The server documents 290 tools with read-only and destructive annotations, OAuth with PKCE and dynamic client registration, and published limits of 60 requests a minute. The single mcp:tools scope grants everything the user can do, sends happen at once with no undo, and no status page, changelog or security contact was found.
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
Fuse MCP
- Keep the client's approval prompt on for every tool that is not read-only. Sends, CRM pushes and Slack posts run at once with no undo
- Launch a campaign only on an explicit request. Creating one makes a draft, and
update_campaign_status_toolstarts real outreach - Call
get_credits_balanceandget_credit_pricesfirst. Every call needs at least 100 credits, and searches cost 2 credits a result - Resolve filter values with
lookup_filter_valuesorget_company_search_enumsbeforesearch_peopleorsearch_companies - Start long work with
deep_research_start, then polldeep_research_status. A tool call has 60 seconds, and runs take 5 to 20 minutes
Retriever
- Don't enrol leads in a sequence with LinkedIn steps without the user knowing LinkedIn forbids automated activity and can restrict their account
- Send an
Idempotency-Keywith each run and replay with the same key after a timeout, so a retry doesn't start a second billed run - Run
getEmail,getPhoneNumberandwebResearchthroughservice_batch_execute, even for one person;service_executerefuses them - Check
creditsbefore a paid run, and callservice_inspectfor a service'sthingsToKnowand credit cost before executing it - 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, Fuse MCP or Retriever?
Fuse MCP scores 46.2 (D) on agent readiness against Retriever's 32 (F), and leads in 4 of 7 scored categories. Retriever leads on reliability, schema & documentation and agent ergonomics.
Do Fuse MCP and Retriever need an API key?
Fuse MCP uses an OAuth sign-in. Retriever takes an API key or an OAuth sign-in.
Can an agent call Fuse MCP and Retriever without installing anything?
Yes. Fuse MCP has a hosted endpoint at https://mcp.fuseai.com and Retriever at https://api.retriever.run/mcp.
Other comparisons with Fuse MCP or Retriever
- Apollo API + MCP vs Fuse MCP
- Apollo API + MCP vs Retriever
- Clay vs Fuse MCP
- Clay vs Retriever
- Coresignal API + MCP vs Fuse MCP
- Coresignal API + MCP vs Retriever
- Crustdata API + MCP vs Fuse MCP
- Crustdata API + MCP vs Retriever
- Dievio vs Fuse MCP
- Dievio vs Retriever
- Enrich Layer API + MCP vs Fuse MCP
- Enrich Layer API + MCP vs Retriever
- Explorium AgentSource vs Fuse MCP
- Explorium AgentSource vs Retriever
- FullEnrich API + MCP vs Fuse MCP
- FullEnrich API + MCP vs Retriever
- Fuse MCP vs Gojiberry
- Fuse MCP vs Hunter API + MCP
- Fuse MCP vs LeadMagic API + MCP
- Fuse MCP vs Lusha API + MCP
- Fuse MCP vs People Data Labs
- Fuse MCP vs Prospeo API + MCP
- Fuse MCP vs RocketReach API
- Fuse MCP vs ZoomInfo API + MCP
- Gojiberry vs Retriever
- LeadMagic API + MCP vs Retriever
- Lusha API + MCP vs Retriever
- People Data Labs vs Retriever
- Prospeo API + MCP vs Retriever
- Retriever vs RocketReach API
- Retriever vs ZoomInfo API + MCP
- Clearbit vs Fuse MCP
- Clearbit vs Retriever
- Databar vs Fuse MCP
- Databar vs Retriever
- Dropcontact API + MCP vs Fuse MCP
- Dropcontact API + MCP vs Retriever
- FullContact vs Fuse MCP
- FullContact vs Retriever
- Retriever vs ZeroBounce
- Hunter API + MCP vs Retriever
- Fuse MCP vs ZeroBounce
Machine-readable
- This page as Markdown
/compare/fuse-mcp-vs-retriever.md· slim.min.md· JSON.json(or sendAccept: text/markdown) - Each listing in full
/api/v1/tools/fuse-mcp.json·/api/v1/tools/retriever.json - From a terminal
anchor compare fuse-mcp retriever(the CLI) - Over MCP
compare_tools {"a": "fuse-mcp", "b": "retriever"}at/mcp, no key