Comparison · verified 2026-10-26
UncensoredAI vs OpenRouter
A very large model marketplace with automatic provider fallbacks.
- Our catalog
- 37
- models across 8 families
- OpenRouter
- 467
- model ids returned by their public /api/v1/models endpoint
- Models both carry
- 4
- compared line by line below
- Cheapest here
- $0.04
- DeepSeek V4.1 Flash · per 1M input
Per-token rates on the same models
Both columns are the provider’s own published rate, normalised to USD per 1,000,000 tokens. Green means the lower price on that column; red means the higher one. Cached input is shown separately because a large repeated prefix can dominate a bill even when the headline input rate looks cheap.
| Model | Input · here | Input · OpenRouter | Cached · here | Cached · OpenRouter | Output · here | Output · OpenRouter |
|---|---|---|---|---|---|---|
| DeepSeek V4.1 Flash 1017deepseek-v4.1-flash-1017 | $0.04 | $0.30 | $0.01 | $0.006 | $0.04 | $1.20 |
| DeepSeek V4 Prodeepseek-v4-pro-0813 | $0.50 | $1.32 | $0.10 | $0.044 | $0.50 | $3.96 |
| GLM-5.3 Flashglm-5.3-flash | $0.056 | $0.15 | $0.0056 | $0.03 | $0.056 | $0.50 |
| Kimi K3.1 Prokimi-k3.1-pro | $0.20 | $2.00 | $0.06 | $0.20 | $0.20 | $10.00 |
Sources · read 2026-10-26
Competitor pricing changes without notice. These figures were read from the pages above on 2026-10-26; re-check before relying on them for a purchasing decision.
What that costs per month
Per-1M rates are hard to budget against, so here is the same DeepSeek V4.1 Flash 1017 priced across four modelled traffic shapes. Token counts are representative of common production patterns, not benchmarks; substitute your own numbers on the calculator.
| Workload | Here | OpenRouter | Difference |
|---|---|---|---|
| Support chatbot100k conversations/month · 500 prompt · 300 completion tokens | $3.40 | $51 | 15.0× cheaper here |
| Coding agent5k agent turns/month · 8k prompt · 1.5k completion tokens | $2.10 | $21 | 10.1× cheaper here |
| RAG answers20k answers/month · 20k prompt with a large cached prefix | $19 | $134 | 6.9× cheaper here |
| Batch classification1M calls/month · 200 prompt · 20 completion tokens | $8.80 | $84 | 9.5× cheaper here |
Commercial terms
The per-token rate is only part of the bill. Credit-purchase fees, minimums and funding methods change the effective cost, so they are listed here rather than buried in a pricing footnote.
- Platform fee
- 5.5% on credit purchases
- Standard plan; the Business plan is 8%. Charged when you buy credits, not per request.
- Free tier
- 50 requests/day
- 25+ free model ids, no card required. Fine for evaluation, not for production volume.
- BYOK ceiling
- $25,000 list price/month, then 5%
- Bring-your-own-key traffic is fee-free below that monthly ceiling.
- Funding
- Card, crypto and more
- Their FAQ states cryptocurrency payments are never refundable.
- Identity
- Email account required
- Needed to hold a credit balance and issue keys.
Where each one wins
A comparison with no downsides is an advertisement. These are the cases where OpenRouter is genuinely the better tool, next to the cases where this gateway is.
Where OpenRouter is the better choice
- 467 model ids, including many open-weight models this catalog does not carry.
- Automatic provider fallback: if one vendor fails, the request re-routes without a code change.
- A real free tier with a daily allowance — the cheapest possible way to evaluate.
- Per-request routing controls, so a provider can be pinned or spend capped per call.
Where UncensoredAI is the better choice
No percentage skim on your balance
OpenRouter charges 5.5% when you buy credits, so $100 of card spend buys $94.50 of inference. Here the balance you top up is the balance you spend; the margin is already inside the per-token rate, which is why input and output prices are lower on every model both services carry. Cached input is the exception: on two of the four shared rows their cache rate is lower than ours, so a workload with a very large repeated prefix can narrow or invert the gap.
Uncensored and abliterated tiers as ordinary ids
Abliterated weights have the refusal direction removed, so the behaviour holds across every request instead of depending on a system prompt. This catalog carries GLM-5.3 Abliterated, DeepSeek V4.1 Flash Abliterated and DeepSeek V4 Pro Uncensored as priced model ids on the same endpoint.
A key is the entire account
No email, password or profile to create. A random 16-digit key is both the login and the API credential, so there is no identity to leak and no signup funnel to abandon.
One rate card instead of a marketplace
Every model here is billed on one rate card with one billing basis. You are not reconciling per-provider spreads, promotional windows or time-of-day price overrides.
The longer version
OpenRouter and this gateway solve different problems, and the honest framing is that OpenRouter is the broader product. With 467 model ids it will almost always carry a model this catalog does not, and automatic fallback means one vendor outage does not take your feature down. If breadth and provider redundancy are the requirement, that is a real advantage and no price table changes it.
Where the two diverge is the shape of the bill. OpenRouter is a marketplace: you buy credits, they take 5.5%, and each request is charged at whatever the routed provider lists, sometimes with time-of-day overrides. This platform publishes one rate card with one billing basis per model. The arithmetic is simpler and, on the models both carry, cheaper — but there is no second provider to fall back to if a model has a bad day.
The uncensored question usually decides it. A marketplace lists what its providers list, and abliterated weights are frequently not among them because the underlying labs do not ship them. This catalog treats abliterated and uncensored variants as first-class model ids priced on the same per-token basis, so switching between the standard and refusal-free build of the same model is a one-string change rather than a different vendor.
Operationally, the difference is what you maintain. OpenRouter gives you routing policy, spend caps and provider preferences to configure. Here there is nothing to configure: one base URL, one key, one catalog. For a small team that is less surface area; for a platform team that wants per-provider control it is less flexibility. Both are legitimate, and which is right depends on whether you are optimising for control or for the shortest path to a working integration.
Questions
- Is UncensoredAI cheaper than OpenRouter?
- On input and output tokens, yes — the table above compares the same model ids at each provider's published rates, and the rate here is lower on every shared row. Cached input is mixed: OpenRouter is cheaper on two of the four shared models, so a workload dominated by a large repeated prefix should check that column before assuming a saving. OpenRouter also charges 5.5% when you purchase credits, which this gateway does not, which widens the gap on a topped-up balance.
- Does OpenRouter have more models?
- Yes, substantially. OpenRouter listed 467 model ids when this page was generated; this catalog carries a curated set across eight families. If you need a specific open-weight model that is not here, OpenRouter is the more likely place to find it.
- Can I use both?
- Yes, and many teams do. Both speak the same Chat Completions wire format, so a client can point at either base URL with an identical request body. A common pattern is this gateway for uncensored and abliterated tiers, and a marketplace for breadth.
- Do I need an account or KYC here?
- No. There is no email, name, phone number or identity check. Generating a key creates the account, and that key is both the login and the API credential.
- What happens if a model has an outage?
- Requests fail with an upstream error and the balance hold is released. Unlike a marketplace with automatic fallback, there is no second provider to re-route to. The live status page probes every model individually so you can see which ones answer before debugging your own code.
Verdict
Choose OpenRouter when breadth and automatic failover matter more than per-token cost, or when you need a model id only they list. Choose this gateway when you want the lowest per-token price on the models it carries, uncensored and abliterated builds as ordinary model ids, and no account to create beyond a key.
Other comparisons
Try the cheaper rate card
Generate a key, point your existing SDK at the base URL, and send a request. There is no email, no card and no KYC step.