uprouter — categories
Unclassified AI API Entries
11 AI API entries awaiting full research — live probe status and risk signals where available. A holding bin, not a ranking or a quality judgment.
- Free tiers & credit values, verified
- Normalized pricing across providers
- Live uptime probes — no pay-to-rank
- Community corrections & reviews
$ uprouter directory --scope unclassified
→ 11 entries · 2 free tiers · $0 in free credits
→ 1 live now · ordered by free-credit value
# a data sort, never sponsored — every row links
# to its full entry with sources + verification dateEvery entry answers to the community
The index is expansive; the community keeps it honest. Votes, reviews and verified corrections feed straight back into rankings and data confidence.
- Vote
Upvote routers that deliver, downvote the ones that don’t — the score re-ranks the index.
- Review
Evidence-based reviews with verification tags. Ratings, not vibes.
- Verify
Corrections that survive review flip the entry to verified — with the evidence on file.
Anyone can edit this index
Every router page has an edit button — propose corrections to descriptions, prices and free tiers. Editors review each proposal against primary sources before it goes live.
- Propose
Open a router page, hit the edit button, and detail your proposed corrections.
- Review
Editors rigorously verify every proposed correction against primary sources.
- Live
Accepted edits are published instantly and logged in the public journal.
// directory
The directory
filtered to unclassified
$ 11 entries match — showing the top 9 · sort: Free credits ↓
- UnclassifiedAPI keydown1693msupdated 20h agocatgpt.ccsource · catgpt.ccoverview
CatGPT is an AI chatbot/router service. Limited public information is available regarding its API documentation or enterprise-grade features. Details were checked against the provider's own page at catgpt.cc.
free quotaDomain serves only a default 'TestPage' placeholder; no active AI API service or free-tier information found on site or in searches.
free models— none trackedsee catgpt.cc’s free quota / risk review → - UnclassifiedAPI keydown10msupdated 20h agoapi.celebras.aisource · celebros.comoverview
Provider slug 'api-celebras-ai' appears to be a typographical error of 'api-cerebras-ai'. No distinct AI service or official documentation was found for this specific domain (celebras.ai). Research suggests 'Celebros' exists as an e-commerce search tool, but it is unrelated to LLM inference.
free quotacelebras.ai does not resolve (DNS fail) - apparent typo of Cerebras, api.cerebras.ai. Cerebras pricing page lists a Free tier; docs have a Free Trial tier (not auto-renewing); trackers report ~1M tokens/day, 5 RPM free.
free models— none trackedsee api.celebras.ai’s free quota / risk review → - UnclassifiedAPI keydegraded329msupdated 20h agocheapyun.cc.cdsource · cheapyun.cc.cdoverview
CheapYun is a legacy AI API relay service. It is currently reported with degraded performance and minimal public documentation. Details were checked against the provider's own page at cheapyun.cc.cd. Some figures in this entry rest on third-party reports rather than a first-party page.
free quotaService down: HTTP 521 (Cloudflare, origin server down) on both HTTP and HTTPS; no active service, hence no free tier. Searches found nothing on this host.
free models— none trackedsee cheapyun.cc.cd’s free quota / risk review → - UnclassifiedAPI keydown683msupdated 20h agocheap-host1.cheapyun.comsource · cheap-host1.cheapyun.comoverview
CheapYun Host1 is a legacy/unclassified endpoint for AI API relaying. It is currently reported as down/degraded and lacks modern documentation. Details were checked against the provider's own page at cheap-host1.cheapyun.com.
free quotaSubdomain serves a '帮助网管理系统' (helpdesk/management system), not an AI API relay; no API pricing or free-tier info found. Likely mis-listed.
free models— none trackedsee cheap-host1.cheapyun.com’s free quota / risk review → - UnclassifiedAPI keyunknownupdated Sep 6, 2026galadriel.comsource · galadriel.comoverview
Galadriel was a decentralized AI inference platform. As of 2026, the service appears to be discontinued, with the primary API endpoint (api.galadriel.ai) no longer resolving. Details were checked against the provider's own page at galadriel.com.
free quotaNo free tier — the source marks this service as deprecated/discontinued.
free models— none trackedsee galadriel.com’s free quota / risk review → - UnclassifiedAPI keydown6001msupdated 20h agosss.jumped.cc.cdsource · sss.jumped.cc.cdoverview
sss.jumped.cc.cd (Jumped) appears to be a private or community-run AI model relay/aggregator. It likely operates on a 'One API' or 'New API' protocol, common for shared access to upstream providers. Details were checked against the provider's own page at sss.jumped.cc.cd.
free quotaSite unreachable at research time: TLS via Cloudflare completes but no HTTP response; no active service observed. Searches found nothing on this host (free .cc.cd subdomain).
free models— none trackedsee sss.jumped.cc.cd’s free quota / risk review → - UnclassifiedAPI keydown241msupdated 20h agolzjapi.cloudsource · lzjapi.cloudoverview
LZJAPI is an obscure AI API provider (lzjapi.cloud) that was unreachable during research in September 2026. No official documentation, pricing, or model lists could be verified from the domain, which appears to be offline.
free quotaDomain does not resolve (DNS 'Name or service not known'); no active service. No free-tier information found and no useful search results.
free models— none trackedsee lzjapi.cloud’s free quota / risk review → - UnclassifiedAPI keydown6001msupdated 20h agomicrosand.aisource · microsand.aioverview
Microsand AI is an AI infrastructure provider that appears to be transitioning or offline as of September 2026. Official domains return errors or lack public-facing documentation, with no verifiable pricing or model availability found.
free quotaSite unreachable at research time: TLS handshake via Cloudflare completes but no HTTP response on ports 80/443; searches found nothing on microsand.ai as an API provider.
free models— none trackedsee microsand.ai’s free quota / risk review → - UnclassifiedAPI keyup771msupdated 20h agoplayhop.comsource · playhop.comoverview
Playhop is a web-based gaming platform offering a variety of casual games. It does not appear to provide a public AI model API or LLM routing service for developers. Details were checked against the provider's own page at playhop.com.
free quotaNot an AI API - Playhop is a free online browser-games portal (20,000+ browser games, no login). Mis-listed in this directory; no AI API or free credits apply.
free models— none trackedsee playhop.com’s free quota / risk review →
// sources & references
Sources & references
Claims on this page are checked against the providers’ own pricing and policy pages (verified 2026-10-12) and against the UPROUTER public dataset.
- Public dataset (CSV, CC BY 4.0) UPROUTER.ONLINE — The {count} rows, probe status, last-verified dates and the fields that remain unquantified
- About & methodology UPROUTER.ONLINE — The research pass, category assignment, and the open editing model this page points corrections to
- Live status board UPROUTER.ONLINE — The probe data behind the status column
- Corrections & contact UPROUTER.ONLINE — The route for category decisions and corrections the edit flow cannot make
- Directory UPROUTER.ONLINE — The same rows in the full, filterable list, alongside the researched categories they are candidates for
// top free allowances
Top free allowances in the Unclassified category
ranked by data, never sponsored
The largest free allowances in the Unclassified category, ranked by normalized free-credit value — a data sort, never sponsored. Values are USD estimates derived from each provider's published signup bonuses, monthly quotas and usage allowances at pay-as-you-go rates; where a figure required an assumption it carries an est. marker.
| # | Provider | Category | Free credits | Access | Status |
|---|---|---|---|---|---|
| 1 | Celebras | Unclassified | not quantified | API key | down· 10 ms |
| 2 | PromptQL (Unofficial/Experimental) | Unclassified | not quantified | Web cookie | unknown |
Free tiers change without notice — confirm current API limits on the provider's official pricing page before you rely on them.
// free vs. paid
Free AI API key vs. paid AI API: which one do you need?
how to choose
A free AI API key is the right starting point for prototypes, hobby projects, model evaluations and low-volume production — you test models and integration paths with $0 of your own spend. A paid API key buys headroom: higher rate limits, priority capacity, SLAs and first access to new models.
| Dimension | Free AI key | Paid AI key | Uprouter directory |
|---|---|---|---|
| Cost | $0 until the allowance runs out | Per-token or per-plan billing | $0 — the directory itself |
| Limits | Quotas, API rate limits, expiry | Vendor SLA & priority capacity | Shows each provider’s current limits |
| Models | Free subset of the catalog | Full catalog | The whole catalog, cross-searchable |
| Failure mode | 429 when the quota is dry | Downtime | Live probes flag dead endpoints |
| Best for | Prototypes, evals, low volume | Production at scale | Choosing which key to use |
The hybrid pattern works well in practice — and it's exactly what this index is for: compare a free allowance against the paid headroom for the same models, check the live status before you commit, and see the break-even point in writing. Every entry records both the free tier and the paid pricing, so the free-versus-paid decision is data, not guesswork.
// one key, every provider
Keep every free AI key working
live status, on every entry
The problem with five free AI keys is not the cost — it's the drift: allowances change, rate limits tighten, and endpoints go quiet without any notice. The index tracks all of it: every entry records the current free tier, its credit value, the paid fallback price, and a live status probe that runs on a schedule so a dead endpoint is flagged before it costs you an afternoon.
That is how a stack of free keys stays useful: you know which allowance expires next, which endpoint just started failing, and what the paid price looks like when you outgrow the free tier — on one screen, with no login required.
$ open /status
→ live uptime board, no account needed
# every endpoint probed on a schedule —
# up / degraded / down with the last-seen time,
# plus a per-entry board on each router page
Recomputed from the live AI provider directory on every request — counts shift as providers change their tiers and pricing.
// tldr
TL;DR
the short version
- A holding bin, not a ranking: 11 published entries whose category, billing and free-tier details have not yet been fully researched and verified. The label describes research state, not quality.
- The current signal is mostly status: at last fetch, 6 of the 11 entries probed down, 2 degraded, 3 up — and risk is unrated on most rows. Read this page as "endpoints the probes touch," with extra caution, not as a list of usable APIs.
- The correction path is the point of the page: any entry here can be researched, sourced and reclassified through the open editing model (/about); category calls route to /contact.
// the unclassified guide
What "unclassified" means in this directory
Every entry in the dataset carries a category — paid-api, official-free-tier, free-relay, subscription-router, and the rest — assigned when research verifies the endpoint: what it is, how it bills, what its free tier is, and where the source is. The 11 entries in this bin are published but past due for that pass: a new endpoint arrived, an entry lost its verification, or the research did not reach a confident category. "Unclassified" is a research-state label, deliberately not a quality judgment — a down probe on an unclassified row and a down probe on a fully researched row are the same kind of fact.
The practical consequence is that this page can say less than the other category pages, and it says so. There is no normalized pricing (the per-1M columns read empty or near-zero), mostly unquantified free tiers, and risk unrated on most rows. What the page can say with confidence is the probe signal — the status column is live data on endpoints that real reachability checks touch — and that is what the table below is for.
How to read (and fix) this list
Read it as a status board with a research queue attached. If an entry here is one you use or trust, the useful action is to research it: confirm the category, the billing unit, and the free-tier terms at the provider's own page, then submit the correction — the open editing model in /about describes exactly how, and /contact is the route for category calls the edit flow cannot make. If an entry here is one you do not recognize, the probes are doing the useful work for you: the down and degraded rows are the ones worth reporting, because a dead endpoint is a data fact anyone can verify, and the bin is where that churn is most visible.
What decides whether an entry leaves this bin
Three facts, in order: what the endpoint is (a model API, a product surface, a relay — the category), how it bills (the unit and the free tier, against the provider's published page), and who operates it (identity, terms, source links). An entry with all three verified moves to its category with the evidence attached; an entry with none stays here with its probe signal visible. There is no minimum time in the bin and no rank within it — the count is a queue, not a leaderboard, and the page says so instead of dressing the queue up as a list.
If you want a row researched
- 1.Pick the entry from the table and note its current probe status and last-verified date — that is the baseline the correction is measured against.
- 2.Find the provider's official page for the endpoint: the pricing or docs page that states what it is, how it bills, and the free-tier terms. If no official page exists, that is itself the finding — the endpoint may be a dead or unofficial surface.
- 3.Submit the research through the open editing model (/about): category, billing unit, free tier, source URL, verification date. The edit lands with the source attached, the way every verified row in the dataset works.
- 4.For calls the edit flow cannot make — merges, removals, category decisions on ambiguous endpoints — route it to /contact with the entry slug and the evidence. The bin shrinks by corrections, not by curation.
// how this page is built
How this list is built
last verified 2026-10-12 — dataset + cited sources checked on this date
predicate: Inclusion: published, non-internal entries whose category is unclassified — the research holding bin. 11 entries matched at last check. This page is intentionally shorter than the other category pages (owner decision recorded 2026-10-12): a bin of 11 low-signal entries cannot support the full editorial template without padding, and the site rule is to state the limit rather than pad.
sort: The shared category sort (free-credit value descending, unquantified last, ties alphabetical) — most rows carry no quantified allowance, so the practical order is alphabetical. The sort is disclosed as-is; it is a queue order, not a ranking.
freshness: Probe status at render time; per-entry last-verified dates; bin membership changes as research passes complete and corrections land.
limitations: No normalized pricing, mostly unrated risk, no operator research for most rows — by definition of the bin. External sources are not cited, stated: the operators here lack a stable, citable documentation set, and a source list would be a patchwork of dead links; the per-entry source links on the entry pages carry whatever exists.
Full methodology, probe cadence and the correction pathway: about & methodology · raw data: public dataset (CC BY 4.0) · pricing and status on every row are measurements with dates, not guarantees.
Written by Mike Fleming — Founder & Editor. He runs the directory, writes the entry research and reviews every page here; corrections go to the contact page.
// faq
Frequently asked questions
Does "unclassified" mean low quality?
No. It means research state: the category, billing and free-tier pass has not been completed or has lapsed. Quality signals — probe status, risk tier where rated, community votes — are shown on the rows exactly as they exist. A down probe here and a down probe on a fully researched row are the same kind of fact.
Why are so many rows down?
The bin skews toward small, new or abandoned endpoints, which is where churn is highest — and the probes surface that churn where it is. At last fetch, 6 of the 11 entries probed down and 2 degraded. A down row is a data fact about reachability, not a verdict on the operator.
Why is risk unrated on most rows?
The risk tier is assigned during the research pass — operator identity, terms, transparency — and rows that have not had that pass carry "unrated" rather than a default low. An unrated row is "not assessed," which the page displays as-is instead of implying safety.
Can I use an API from this list?
The probe tells you whether the endpoint answers; nothing else here is verified — billing, terms, operator and free-tier details all await the research pass. Before any real use, verify those at the provider's own page. That is the same standard the other category pages meet automatically; here you would be meeting it manually.
Why is there no pricing on these entries?
The per-1M normalization requires a published rate table, and rows without a verified one show empty rather than an estimate. That is the no-invented-numbers rule applied to a category where most operators publish nothing citable — the blank is the honest value.
How does an entry get into this bin in the first place?
Publication without a completed research pass: a new endpoint landed, a verification lapsed, or a category call was left open. The bin is the default state for unverified entries, not a penalty — which is also why the correction path is the main content of this page.
Is this page ranked?
No. The disclosed sort is the shared category sort (credit value descending, unquantified last, ties alphabetical), but most rows carry no quantified credit, so the practical order is alphabetical. The count is a queue, and the page refuses to present it as a leaderboard.
What happens to an entry when it is reclassified?
It moves to its category with the evidence attached, this page re-renders without it, and the change is visible in the entry's history. The /s/<slug> URL does not move — the entry stays where it is while the category and the table update.
from the blog
Notes on free AI keys, tokens and credits
Free Tiers That Changed in 2026: Six Verified Pricing Shifts
GitHub Models was retired, Perplexity's Sonar API sunsets September 27, Cerebras's starting credit is now $5, and OpenAI dated its promotional pricing. Here is what changed and what to do.
Free Tiers for AI Agents: Your Token Budget Is a Request Budget
An agent does not make requests the way a chat user does — one turn can be five round-trips, each re-sending the whole conversation. Free tiers that look generous on tokens are brutal on agents.
Daily vs Monthly vs One-Time: Free Tier Reset Shapes Compared
Free tiers reset on three different clocks — daily pools, monthly credits, and one-time grants that never refill. The clock, not the dollar value, decides which shape fits a workload, and the best answer is a portfolio.
// by category & list
Browse by category & list
12 live pages
categories — one page each, root-level URLs
data-ranked lists — ordered by data, never sponsored
Every page carries its own live totals (routers, free tiers, free credits, live now) and its own filtered directory — generated from the same dataset as the rest of the index.