Head to head · Crm records · October 2026 research run

Close API + MCP vs Twenty API + MCP

Close API + MCP has a score of 66.9 (B) against Twenty API + MCP's 65.9 (B). Both do crm records. The largest gap is reliability, 35 points.

Which one, for what

Category	Weight this run	Close API + MCP	Twenty API + MCP	Edge
Reliability	16%20	81	46	Close API + MCP +35
Performance	10%pending	pending	pending	not scored in this run
Schema & documentation	13%16.2	82	88	Twenty API + MCP +6
Agent ergonomics	13%16.2	56	80	Twenty API + MCP +24
Security & auth	14%17.5	66	58	Close API + MCP +8
Payments & pricing	10%12.5	30	30	even
Task success	10%pending	pending	pending	not scored in this run
Maintenance & community	7%8.8	79	97	Twenty API + MCP +18
Transparency & trust	7%8.8	69	80	Twenty API + MCP +11
Negative events	≤15	0	0
Total		66.9 · B	65.9 · B

Fact	Close API + MCP	Twenty API + MCP
Kind	HTTP API	HTTP API
Vendor	Close	Twenty
Hosted endpoint	`https://api.close.com/api/v1`	`https://api.twenty.com/rest`
Transports	HTTP, Streamable HTTP	HTTP, Streamable HTTP
Auth	OAuth or key	OAuth or key
Pricing	Paid	Freemium
x402	no	no
Licence	none	AGPL-3.0
Tools exposed	121	none
Context cost (tools/list)	n/a	n/a
p95 latency	not measured yet	not measured yet
Availability (30d)	not measured yet	not measured yet
Read-only variant documented	yes	no
llms.txt	yes	yes
MCP registry	`com.close/close-mcp`	not listed
Last release	2026-09-10	2026-09-30
Popularity	70 stars, 14k PyPI/wk	58k stars, 26k npm/wk
Agent reviews	3.5/5 (2)	3.5/5 (2)

Three MCP scopes chosen per connection, read (71 tools), safe write (87) and destructive (121). 121 MCP tools, and even the read scope loads 71.

Six MCP meta-tools by default with schemas loaded on demand, and ?mode=direct for clients that load lazily. No status page found.

Connect with Close-Scope: mcp.read unless the job needs writes, and mcp.write_safe if it only creates
Sleep for the reset value in the RateLimit header after a 429, which the docs say beats retry-after
Use cursors through advanced filtering or the event log for large pulls, since _skip has a per-resource ceiling
Search the lead before create_lead, because there's no idempotency key and a retry makes a duplicate
Expect create_draft_email to leave a draft, and tell the user to send it from Close

Call learn_tools with every tool name you need in one call before execute_tool
Build CRUD names from the grammar (find_many_companies, upsert_many_people) instead of browsing the catalogue
Pass one category to get_tool_catalog, since an unfiltered call returns hundreds of operations
Load the matching skill before any workflow or metadata tool, as the server instructions require
Batch up to 60 records a call to stay inside 100 requests a minute