Live demo. This is Limelit Open running unmodified, at commit
83bddd036688.
It is read-only for visitors.
Run your own →
NorthWindnorthwind.co
last run 2026-09-14 21:48:28
Overview
How NorthWind shows up when people ask AI about your category.
Visibility
?The share of answers that name you. Branded prompts are left out, and an answer surface that did not render is left out entirely.
67%
746 of 1,106 answersfirst window6 days, 56 to 71%
Share of voice
?Your mentions divided by all tracked brands' mentions in the same answers. One answer naming you twice counts twice.
28%
of every tracked brand mention
Average position
?The mean rank of your brand in numbered or bulleted lists. Mentions in prose have no rank and are not averaged in.
2.8
where you appear in a ranked list
Citation share
?The share of all cited URLs that point at your own domains.
12%
of cited sources are yours
The race
7 of 22 brands · 6 measured days · 1,106 answers
How often each brand is named, by day. Values at the right edge are the last measured day. Hollow dots rest on fewer than 20 answers. Shaded spans had no run. The 15 brands below the top seven are in the table further down.
NorthWind you71%
NVIDIA 60%
vCluster 30%
CoreWeave 35%
Lambda 25%
Nebius 20%
Mirantis 16%
Share of voice
Of every time a tracked brand was named in this window, how much was you.
NorthWind28%
NVIDIA17%
vCluster8.7%
CoreWeave7.6%
Lambda6.9%
Nebius5.7%
Mirantis3.7%
15 other brands21%
vMetal3%
Red Hat2.8%
RunPod2.8%
Crusoe2.6%
OpenNebula2%
Nscale1.9%
DigitalOcean1.5%
Northflank1.4%
Spectro Cloud1.1%
Spheron0.9%
WhiteFiber0.5%
Introl0.4%
WWT0.3%
BuzzHPC0.2%
NeuReality0.2%
By engine
Where each brand stands on each engine. Your dot is the large one; the number on the right is your share of that engine's answers. Grey dots are the brands outside the top seven. Rows run from your strongest engine to your weakest.
Where you stand
22 brands
Brand
Seen in
Share
Avg position
1
NorthWind you4498 in 746 answers
67%
28%
2.8
2
NVIDIA2772 in 716 answers
65%
17%
3.0
3
vCluster1381 in 300 answers
27%
8.7%
2.9
4
CoreWeave1200 in 332 answers
30%
7.6%
2.9
5
Lambda1097 in 300 answers
27%
6.9%
3.1
6
Nebius904 in 236 answers
21%
5.7%
3.2
7
Mirantis593 in 179 answers
16%
3.7%
3.2
8
vMetal470 in 174 answers
16%
3%
2.5
9
Red Hat445 in 147 answers
13%
2.8%
2.9
10
RunPod442 in 142 answers
13%
2.8%
3.0
11
Crusoe416 in 102 answers
9.2%
2.6%
3.0
12
OpenNebula310 in 94 answers
8.5%
2%
3.0
13
Nscale295 in 86 answers
7.8%
1.9%
3.0
14
DigitalOcean242 in 109 answers
9.9%
1.5%
3.4
15
Northflank215 in 89 answers
8.1%
1.4%
3.3
16
Spectro Cloud175 in 72 answers
6.5%
1.1%
3.3
17
Spheron137 in 77 answers
7%
0.9%
3.9
18
WhiteFiber73 in 31 answers
2.8%
0.5%
2.8
19
Introl69 in 27 answers
2.4%
0.4%
3.2
20
WWT42 in 23 answers
2.1%
0.3%
3.5
21
BuzzHPC38 in 27 answers
2.4%
0.2%
3.5
22
NeuReality36 in 18 answers
1.6%
0.2%
3.7
Every prompt, every engine
47 prompts × 6 engines
Worst first. Click any cell to read the answers behind it.
Each cell is how many of that engine's answers named you, over how many were asked.
A dash means that prompt has not been asked of that engine.
Who gets cited
Each bar is one day's citations, split by what kind of site they point at. The number on top is your share; the number under the date is how many citations that day had.
own1452 · 12%
competitor3804 · 32%
informational151 · 1.3%
social1038 · 8.7%
other5428 · 46%
What the engines searched for
The words in the searches engines ran on the way to their answers. Bigger is more often. This is often not the vocabulary you used.
None of the engines reported their searches yet.
By question type
A question naming a rival lists rivals by construction. Discovery is the one that measures unprompted reach.