Because faymas is a brand word, not a topic, and Google hands a brand word to whichever domain it holds the most evidence for. On 4 October 2026 the evidence was lopsided in a way you can count: faymas.in listed 13,256 URLs in /sitemap.xml, faymas.net listed 41. Every one of the 50 prompt pages we sampled on faymas.in carries the word in its <title>, against 36 of faymas.net’s 41 pages. The estimated values differ by about 190×. The number of pages that can carry the word differs by about 323×. Those two ratios are not the same number, and we are not claiming one causes the other — but they are the same order of magnitude, and everything else we measured is not.
This guide analyses published prompt text. No generation tests were run.
Both sitemaps fetched 4 October 2026, 10:05–10:25 UTC+8. All 41 faymas.net URLs fetched individually; 73 faymas.in URLs fetched, including a seeded random sample of 50 /prompt/ pages out of 13,198.
The two numbers, and how much weight they can carry
The 905,000 and 4,750 figures are third-party SEO-tool estimates of monthly traffic value. Neither site publishes traffic, and we could not verify either number against any official source. Worse, two third-party sources disagree about the size of the prize itself: one puts faymas at roughly 80,000 monthly searches, another puts it at 1,300. That is a 60× disagreement on the same keyword.
So treat the absolute numbers as unverified. What we could do without anyone’s permission is measure the two sites directly. That is the rest of this page.
What we measured
| Measurement (4 Oct 2026) | faymas.in | faymas.net |
|---|---|---|
URLs in /sitemap.xml | 13,256 | 41 |
Of those, /prompt/ pages | 13,198 | 0 |
| Of those, model landing pages | 0 | 23 |
Pages with “faymas” in the <title> | 50/50 prompt pages sampled; 9/9 category pages | 36/41 |
| Titles that begin with “Faymas” | 2 (homepage and one collection page) | 2 |
| Median body word count | 974 (n=50 prompt pages) | 732 (n=41, all pages) |
Distinct <lastmod> dates in sitemap | 23 | 1 |
| Sitemap count change since 30 Sep 2026 | +1,385 | 0 (41 → 41) |
1. The page gap is wider than the value gap
This is the finding that reframes the question. If you only look at the traffic-value estimates, you see a 190-to-1 contest and go looking for a single cause. If you count the pages each site offers to a crawler, you get 323-to-1 — a bigger gap, in the opposite direction from what a neat explanation would want.
Narrow it to pages that actually put the word in the title tag and the gap widens again. All 50 sampled prompt pages on faymas.in end with | Faymas. So do all 9 category pages and 9 of its 12 collection pages. If the sample holds across the full sitemap, that is roughly 13,217 brand-bearing title tags against 36 on faymas.net — about 367 to 1.
We sampled 0.4% of the prompt pages, so that extrapolation is an inference, not a measurement. But even the measured floor — 13,198 URLs against 41 — is enough to say the two domains are not playing the same game.
2. One sitemap is compounding. The other has not moved.
We have counted faymas.in’s sitemap three times on this project: 11,813 prompt pages on 30 September, 11,965 on 1 October, 13,198 on 4 October. That is +1,233 pages in three days, and the rate looks like it is accelerating rather than settling.
faymas.net’s sitemap held 41 URLs on 30 September. It holds 41 URLs today. Same 41, same order, no additions.
Nothing about this is a ranking factor on its own. It is a difference in what the two sites are. One is a marketplace that grows by one page per upload. The other is a brochure for a generator.
3. On faymas.net the word is a suffix, not a subject
Break faymas.net’s 41 URLs apart and 23 of them are landing pages for somebody else’s model: Veo 3.1, Kling 3.0, Kling 2.6 Pro, Wan 2.5/2.6/2.7, Seedance 1.0/1.5/2.0/2.0 Mini, Nano Banana 2 / 2 Lite / Pro, Seedream 4.5 / 5.0 Lite / 5.0 Pro, GPT Image 1.5 and 2, Gemini Omni, Grok Imagine 1.5, HappyHorse 1.0 and 1.1.
Read their titles in a row and the pattern is flat. Veo 3.1 AI Video Generator | Faymas AI. Kling 3.0 AI Video Generator | Faymas AI. Wan 2.7 AI Video Generator | Faymas AI. Twenty-nine of the 41 titles end with the brand; only 2 begin with it. The subject of the page is the model. The brand is the signature.
Only one URL on the whole site carries the word in its slug — /prompts/faymas-ai-prompts. Add the homepage and that is the entire head-term surface: two pages.
faymas.in is not more elegant. Its 13,198 prompt pages also relegate the brand to a suffix. The difference is arithmetic: a suffix repeated 13,198 times is a very different object from a suffix repeated 36 times.
4. Neither site is publishing thin pages
Worth stating, because it is the obvious explanation and it fails. We fetched all 41 faymas.net pages and read them. Median body length is 732 words; the shortest is 94, the longest 1,338. All 41 returned HTTP 200. All 41 titles are unique.
The 12 /prompts/ collection pages are genuinely different from each other — 12 distinct text hashes, pairwise word-set overlap between 0.16 and 0.47. That matters: on faymas.in we found the equivalent pages to be eleven copies of one grid (measured on 2 October). faymas.net does not have that problem. Its pages are individually respectable.
faymas.in’s sampled prompt pages run longer, median 974 words, range 688 to 2,025. One of the 50 had a duplicate title. So the gap is not “one site writes and the other does not.” It is scale, not craft.
5. The freshness signal sits on one side
Every one of faymas.net’s 41 sitemap entries carries the identical <lastmod>: 2026-10-04T02:00:53.258Z. One timestamp, to the millisecond, across the whole file. That is a build stamp, not a modification date, and it tells a crawler nothing about which pages changed.
faymas.in’s sitemap carries 23 distinct dates, from 2026-07-05 to today, with 13,207 entries stamped 4 October. Whether that stamp reflects real edits or another build process we cannot tell from outside. But the two files make different claims, and only one of them makes a claim at all.
What the link data says, and its unit problem
Third-party backlink data puts faymas.in at domain rating 16 with 5,739 backlinks from 43 referring domains, 100 of them dofollow. It puts faymas.net at domain rating 0 with 754 backlinks from 342 referring domains, 5 dofollow.
Those figures are not cleanly comparable and we will not divide them — the source mixes total links, referring domains and dofollow links in one sentence. But two things survive the ambiguity. One domain has 20× the dofollow links. And faymas.net’s profile is inverted: 342 domains point at it, yet only 5 links pass equity. Roughly 98% of the domains that mention it do not vouch for it.
A third domain, faymas-in.com, is in the same keyword auction at an estimated value of 30 for faymas. We covered its separate role in the three-domain comparison.
What none of this proves
- The value numbers. Third-party estimates, unverified, built on a search-volume figure that a second source disputes by 60×. We could not check them.
- Causation. A 323-to-1 page ratio sitting next to a 190-to-1 value ratio is a correlation we noticed. Google has published no rule saying more pages means more brand-query traffic.
- Indexation. A URL in a sitemap is a URL offered, not a URL indexed. We have no way to count what Google actually holds for either domain.
- The 13,217 figure. Extrapolated from a 50-page sample. The measured floor is 13,198 URLs; the brand-in-title rate above that is an inference.
- Which site deserves to win. The estimates measure search capture, not product quality. faymas.net generates images and video; faymas.in distributes prompt text. They are not the same product.
What to do with it
If you found this page while deciding where to spend your own effort, the useful conclusion is not “one site is bigger.” It is that the two domains are not competing for the same thing, and the keyword overlap is a symptom of that rather than the contest itself.
faymas.in is building a corpus: 13,198 pages, ~400 new ones a day, each with a describable subject and a brand suffix. faymas.net is building a product page set: 41 pages, unchanged in four days, each about a model it rents compute for. If your goal is to have a page of yours found, the first site has 13,198 slots and the second has 41. If your goal is to generate an image, the ranking contest is irrelevant — only one of them has a generator.
And if your goal is to write about faymas: the head term is not up for grabs by publishing better pages. It is already resolved, by page count and by links, in favour of the domain that treated the word as inventory. We measured 323-to-1 on 4 October 2026. We will not guess what it is next month — the one growing side is the one that decides it.