MCP Sequential Thinking server 1.0.0 (Node 20) / 2.0.0 (Node 24): our test log

MCP Sequential Thinking server answered over stdio in 271 ms cold (v1.0.0, Node 20) and 159 ms (v2.0.0, Node 24); median call 0.8 to 0.4 ms.

By Agentic Weekly, operated independently from Texas, US · Published Oct 25, 2026 · Last updated Oct 26, 2026 (America/Chicago) · How we work

Tested by us on 2026-10-25. Local open-weight model, not the vendor's hosted models. A smoke test on a CPU-only box, not a benchmark.

Result at a glance

MCP Sequential Thinking server test record
Field	Value
Result	Passed
Version tested	1.0.0 (Node 20) / 2.0.0 (Node 24)
Date and time	2026-10-25, 22:37 to 23:08 CDT (America/Chicago)
Task	A small Python client written with the official MCP Python SDK (2.3.0) spawns the server as a child process over stdio, times the initialize handshake (cold start: spawn to initialize result), times one tools/list, then times 30 sequential tools/call requests and reports the median and 95th percentile. Numbers are milliseconds on the rig described on the methodology page.
Model used	none (no language model involved)
Hardware	Linux x86_64 virtual machine (Debian 13), 8 vCPU (Intel Xeon, 1 thread per core), 15 GiB RAM, no GPU
Node 20 / sequential-thinking-server 1.0.0	cold start 271 ms; tools/list 6.6 ms (1 tools); 30 x sequentialthinking: median 0.8 ms, p95 1.5 ms
Node 24 / sequential-thinking-server 2.0.0	cold start 159 ms; tools/list 5.9 ms (1 tools); 30 x sequentialthinking: median 0.4 ms, p95 1.4 ms

What we found

On Node 20 npm resolved version 1.0.0; on Node 24 it resolved 2.0.0. The 2.x line needs Node 24 or newer, and npm picks an older version on Node 20 without telling you.
Cold start fell from 271 ms (1.x, Node 20) to 159 ms (2.x, Node 24), and the median call from 0.8 ms to 0.4 ms; both are well under a frame of video, so the practical difference is the startup, not the calls.
The server exposed 1 tools on 1.x and 1 on 2.x.

What we ran, step by step

Commands run, wall-clock time and exit code
Step	Wall time	Notes
npm install on Node 20.19.2 (cold)	2.8 s	Resolved 1.0.0; 30 MB in 130 packages; load 3.58.
npm install on Node 24.22.1 (cold)	0.9 s	Resolved 2.0.0; 18 MB in 31 packages; load 1.73.
Latency run: Node 20 / sequential-thinking-server 1.0.0	n/a	sequential-thinking-server 1.0.0 (protocol 2025-11-25). cold start 271 ms; tools/list 6.6 ms (1 tools); 30 x sequentialthinking: median 0.8 ms, p95 1.5 ms. Load 2.50.
Latency run: Node 24 / sequential-thinking-server 2.0.0	n/a	sequential-thinking-server 2.0.0 (protocol 2025-11-25). cold start 159 ms; tools/list 5.9 ms (1 tools); 30 x sequentialthinking: median 0.4 ms, p95 1.4 ms. Load 1.45.

Exact commands (paths shortened)

# npm install on Node 20.19.2 (cold)
npm install --no-audit --no-fund --cache /tmp/npmcache @modelcontextprotocol/server-sequential-thinking

# npm install on Node 24.22.1 (cold)
npm install --no-audit --no-fund --cache /tmp/npmcache @modelcontextprotocol/server-sequential-thinking

# Latency run: Node 20 / sequential-thinking-server 1.0.0
node work/npm/mcp-server-sequential-thinking/node_modules/.bin/mcp-server-sequential-thinking

# Latency run: Node 24 / sequential-thinking-server 2.0.0
/workspace/tests2/tools/node-v24.22.1-linux-x64/bin/node work/npm24/mcp-server-sequential-thinking/node_modules/.bin/mcp-server-sequential-thinking

Raw output (excerpts)

Output of the timed call (Node 20 / sequential-thinking-server 1.0.0), first 300 characters

sequentialthinking({"thought": "t", "nextThoughtNeeded": false, "thoughtNumber": 1, "totalThoughts": 1})
{
  "thoughtNumber": 1,
  "totalThoughts": 1,
  "nextThoughtNeeded": false,
  "branches": [],
  "thoughtHistoryLength": 30
}

Gotchas and things that broke

If a tutorial says 'npx -y @modelcontextprotocol/server-sequential-thinking' and you are on Node 20 or 22, you will not get the latest 2.x server; check the version string the server reports in its initialize result.

What we did not test

Remote (HTTP) transport, authentication and concurrent clients: only local stdio with one client was measured.
Large payloads: calls used tiny inputs, so these latencies are protocol and process overhead, not throughput.
Anything about answer quality: no model was involved in the latency numbers.
These are single smoke-test runs; it is not a benchmark. The machine was shared with other jobs.

Reproduce it

Versions change. These commands install whatever is latest on the day you run them unless a version is pinned, so expect different results later. The version we measured is in the table above.

npm install @modelcontextprotocol/server-sequential-thinking
# then drive it with any MCP client over stdio; we used the Python SDK client in the research page's script description
node node_modules/.bin/mcp-server-sequential-thinking

How to read this test

Every number on this page was measured on this one machine, which also ran other jobs, so timings are indicative and vary run to run. A pass means our small task worked once or as stated, a fail means it did not work as we ran it; neither says how the tool behaves with a vendor's frontier model, on a GPU, or at scale. Where a tool could not run without an account or a GPU, we say so instead of guessing.

See the testing rig for the hardware, the local model and the rules we follow, and the test hub for every other test.

MCP Time server: The MCP Time server 2026.10.19 started in 483 ms and answered get_current_time in 0.8 ms median.
Notion MCP server: Notion MCP server 2.5.2 (@notionhq/notion-mcp-server) started over stdio in a median 316 ms and listed 24 tools in a median 16 ms; we did not make any authenticated call.
Playwright MCP: Playwright MCP 0.0.83 exposed 25 tools and navigated a headless Chrome to a local page in 40.5 ms median (cold start 383 ms).
Sentry MCP server: Sentry MCP server 0.42.0 (@sentry/mcp-server) started over stdio in a median 1134 ms and listed 9 tools in a median 13 ms; we did not make any authenticated call.
Supabase MCP server: Supabase MCP server 0.14.0 (@supabase/mcp-server-supabase) started over stdio in a median 468 ms and listed 19 tools in a median 363 ms; we did not make any authenticated call.
Chrome DevTools MCP: Chrome DevTools MCP 1.12.0 exposed 32 tools and opened a page in headless Chrome in 136.6 ms median; our first navigate_page call was rejected for a missing pageId.

All test logs · How we test

FAQ

How fast is the MCP Sequential Thinking server?

On our shared 8 vCPU VM: 271 ms to start and answer initialize on Node 20 (v1.0.0) and 159 ms on Node 24 (v2.0.0); a call took 0.4 ms at the median.

Sources

MCP reference servers repository (fetched 2026-10-25)

Markdown version for agents: /tests/mcp-server-sequential-thinking.md