n8n 2.44.6: our test log

n8n 2.44.6 installed from npm on Node 24 in 1 min 54 s (2,343 packages), imported a three-node workflow from the CLI and executed it: a Code node returned 42 and a request to a local model returned "42".

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

n8n test record
Field	Value
Result	Passed
Version tested	2.44.6
Date and time	2026-10-25, 22:45 to 23:00 CDT (America/Chicago)
Task	Install n8n from npm, import a workflow with a Manual trigger, a Code node that returns {a:17, b:25, sum:42} and an HTTP Request node that asks the local Qwen2.5-3B model "What is 17 + 25?", and execute it headless with `n8n execute`. Pass means status success with sum 42 and the model answer containing 42.
Model used	Qwen2.5-3B-Instruct (Q4_K_M) through Ollama 0.40.11
Hardware	Linux x86_64 virtual machine (Debian 13), 8 vCPU (Intel Xeon, 1 thread per core), 15 GiB RAM, no GPU
Install (npm, Node 24)	1 min 54 s, 2,343 packages
Execute (CLI, incl. start)	3.5 s
Workflow status	success

What we found

The CLI execution finished with status success: the Code node produced the sum 42 and the Ollama-compatible request returned the text 42.
n8n logged a warning that it could not start its Python task runner because its virtual environment is missing; the JavaScript Code node (what we used) worked. Python Code nodes were not tested.
The log also lists deprecations: running outside a container is deprecated and internal task-runner mode is going away, so the supported path is shifting toward Docker.

What we ran, step by step

Commands run, wall-clock time and exit code
Step	Wall time	Notes
npm install on Node 24	1 min 54 s	added 2,343 packages in 2m (user 1m36s, sys 28s); real 1m54s.
Import the workflow	n/a	Needed an explicit workflow id (the import file sets "id": "smoke1").
Execute	3.5 s	real 0m3.481s for the CLI process including start; the workflow itself ran about 1 s (Code node 504 ms the first time, model request about 0.74 s in the later run).

Exact commands (paths shortened)

# npm install on Node 24
npm install n8n

# Import the workflow
n8n import:workflow --input=wf-import.json

# Execute
n8n execute --id=smoke1

Raw output (excerpts)

wf.json (the whole workflow)

{"name":"smoke","nodes":[
{"parameters":{},"name":"Manual","type":"n8n-nodes-base.manualTrigger","typeVersion":1,"position":[0,0]},
{"parameters":{"jsCode":"return [{json:{a:17,b:25,sum:17+25}}];"},"name":"Code","type":"n8n-nodes-base.code","typeVersion":2,"position":[220,0]},
{"parameters":{"method":"POST","url":"http://127.0.0.1:11434/v1/chat/completions","sendBody":true,"specifyBody":"json","jsonBody":"={{ JSON.stringify({model:'qwen2.5-3b',temperature:0,max_tokens:20,messages:[{role:'user',content:'What is '+$json.a+' + '+$json.b+'? Answer with just the number.'}]}) }}","options":{}},"name":"Ollama","type":"n8n-nodes-base.httpRequest","typeVersion":4.2,"position":[440,0]}
],"connections":{"Manual":{"main":[[{"node":"Code","type":"main","index":0}]]},"Code":{"main":[[{"node":"Ollama","type":"main","index":0}]]}},"settings":{}}

Execution output (first lines, trimmed)

n8n Task Broker ready on 127.0.0.1, port 5679
Failed to start Python task runner in internal mode. because its virtual environment is missing from this system. Launching a Python runner in internal mode is intended only for debugging and is not recommended for production. Users are encouraged to deploy in external mode. See: https://docs.n8n.io/hosting/configuration/task-runners/#setting-up-external-mode
[license SDK] Skipping renewal on init: renewOnInit is disabled in config
[license SDK] Skipping renewal on init: license cert is not initialized
Registered runner "JS Task Runner" (77CEduUM-yHY2wluYh1kZ) 
Execution was successful:
====================================
{
  "data": {
    "vers

Install log

Sun Oct 25 22:45:55 CDT 2026

added 2343 packages in 2m

real	1m54.151s
user	1m35.955s
sys	0m27.970s
EXIT=0
Sun Oct 25 22:47:49 CDT 2026

Gotchas and things that broke

This run used Node 24. An earlier note in our repository recorded n8n failing to run on Node 20; we did not re-run Node 20 for this log, so we make no claim about it.
2,343 packages is a large dependency tree; install took nearly two minutes even on a fast link.

What we did not test

The editor UI, webhooks, credentials, scheduled triggers, community nodes, Python Code nodes and queue mode.
Docker deployment, which n8n now recommends.

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.

# Node 24
npm install n8n
npx n8n import:workflow --input=wf-import.json
npx n8n execute --id=smoke1

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.

Open WebUI: Open WebUI 0.12.4 installed with uv in 11.6 s (6,977 MB in the virtualenv), was healthy about 11 to 12 s after start, and answered a chat through its API in 0.68 s using a local Ollama model.
Flowise: Flowise failed to install from npm on Node 24.22.1 after 106.2 s: the better-sqlite3 native build needed a C++ toolchain this machine lacks.
Langflow: Langflow 1.13.0 installed in 19.76 s (1.7 GB) and its backend was healthy 91.5 s after launch, serving 26 starter flows and 416 components.

All test logs · How we test

FAQ

Can n8n call a local Ollama model?

Yes: an HTTP Request node posted to the OpenAI-compatible /v1/chat/completions endpoint of a local Ollama and returned 42 in our run. We did not use the dedicated Ollama node.

Sources

n8n repository (fetched 2026-10-25)

Markdown version for agents: /tests/n8n.md