---
title: "LoreBary proxy setup: your own key, uncensored model"
canonical: https://uncensoredllm.ai/guides/lorebary-proxy
generated: 2026-11-09T16:02:05.951Z
source: https://uncensoredllm.ai/guides/lorebary-proxy
base_url: https://api.uncensoredllm.ai/v1
---

# Sophia's LoreBary proxy with an uncensored model

What LoreBary's Proxy Portal passes on (a proxy URL, your key, a model id), which of its features survive a custom proxy, and the direct route when you only want the model.

Steps checked November 6, 2026 · prices as of November 9, 2026

## What LoreBary is, and where this key fits

[Sophia's LoreBary](https://lorebary.com) is a roleplay platform with a proxy layer in front of it. Its Proxy Portal, read 2026-11-02, says what it does in one line: "Use our models and any provider key inside the chat app you already have. Three fields, no lock-in." You keep Janitor AI, Chub or Wyvern, point them at a LoreBary URL, and LoreBary adds its own layer on the way through: server commands typed mid-chat, plugins with dice and stats, lorebooks, personas, auto summary. The portal says it adds no fee; you bring the model.

That makes it a different kind of page from the other client guides here. LoreBary is not where the model is set; it is a hop between your front-end and the API. Two routes, and the right one depends on what you want:

- **You want the model, not the layer.** Put the API straight into the front-end. Janitor AI, Chub, Wyvern and RisuAI each take the same three values, and the [Janitor AI](https://relayswitch.io/guides/janitorai), [Chub AI](https://uncensoredllm.ai/guides/chub-ai), [Wyvern](https://uncensoredllm.ai/guides/wyvern-chat-api) and [RisuAI](https://uncensoredllm.ai/guides/risuai-custom-api) guides walk the fields. Nothing in the middle, one bill, every request on the console's Usage page.
- **You want LoreBary's layer on top.** Then the key goes through LoreBary, and the rest of this page is about what that passes on, what it keeps, and what it costs.

- Proxy URL: the URL LoreBary gives you For a direct setup it is https://api.uncensoredllm.ai/v1 (or https://api.uncensoredllm.ai/v1/chat/completions where the app posts to the address as typed)
- API key: sk-ullm-live-… Created in the console, shown once; LoreBary passes it on
- Model: uncensored-ai-roleplay Any id from the Models page, typed exactly

## Prerequisites

- A LoreBary account, and the front-end you already use signed in in the same browser. The Proxy Portal is at [lorebary.com/connect-proxy](https://lorebary.com/connect-proxy).
- A key. Create one in the [console](https://app.uncensoredllm.ai/console). The first top-up is $10, there is no subscription, and bonus credit is added on larger top-ups (+5% at $200, +10% at $500, +15% at $1,000; these are the tiers the console applies). The key starts with `sk-ullm-` and is shown once. Make it a key of its own, with a monthly limit, because it will sit in a third party's configuration.
- A model id from the [Models page](https://uncensoredllm.ai/models). Three starting points are priced below.

## The Proxy Portal's two paths

The portal asks "How do you want to use LoreBary?" and offers two cards. **Connect my current app** ("You already have your chats, cards and habits somewhere, and a provider key you pay for") opens with **Show me the setup**, walks you to **Your own key**, and lists the providers LoreBary has a permanent route for. For each it hands you three values to paste into the front-end: the **Proxy URL** on `api.lorebary.com`, your **API key**, and the **Model** id "including any suffix".

This API was not on that list on 2026-11-02. So the path for this key is the second card, **My Custom Proxy** (the older community guide at lorebary-guide.carrd.co, read 2026-11-02 and marked outdated on its own page, calls the same form "Cheating On Us?"). The form, by what each field asks for:

1. **Endpoint URL**: `https://api.uncensoredllm.ai/v1`. The older guide says to paste the provider's curl command or endpoint here; the base URL with `/v1` is the endpoint.
2. **Path to append**: the chat completions path, `/chat/completions`, where the form offers a choice.
3. **API type**: **OpenAI**. The form also offers Gemini and Anthropic; every chat model on this site answers the OpenAI format.
4. **Create Proxy**. LoreBary answers with a proxy URL of its own. That URL, not `https://api.uncensoredllm.ai/v1`, is what goes into the front-end's proxy field; the key and the model id go into their own fields as usual.

In Janitor AI that is Settings → Provider **Proxy** → **+ New**: the LoreBary URL as the proxy URL, your key, the model id, **Add**, **Save**, then reload the page, as the older guide says to. Per rpwithai.com's LoreBary walkthroughs (published 2025-12-15 and 2026-02-01), Chub wants a LoreBary URL with no `/v1` on it and Wyvern wants it with `/v1`.

The form sits behind a LoreBary sign-in, and the labels above are the older guide's; LoreBary renames things between releases. Match each field by what it asks for: a URL, a path, a type.

## What a custom proxy keeps, and what it loses

The honest part. The older community guide (lorebary-guide.carrd.co, read 2026-11-02, marked outdated on its own page) says LoreBary's features (commands, plugins, lorebooks) work only when the proxy URL belongs to one of its permanent provider routes. A custom proxy relays the chat and nothing else. So with this key behind a custom proxy you get: your front-end, the model you named, and the LoreBary account for its library and forums. You do not get the mid-chat commands or the plugin mechanics on that route.

Which is why the first route on this page exists. If the layer is the reason you are here, the thing to ask LoreBary for is a permanent route to `api.uncensoredllm.ai`; its Discord is where that request goes. If the model is the reason, skip the hop: the direct guides take two minutes and nothing is lost on the way.

## Which model to name

The ids LoreBary's provider list prints for other routes (`deepseek-chat` and the like) are not on this list. Three starting points from this catalog, each an id the Model field takes as written, priced live:

| Model | Build | Input $/M | Output $/M |
| --- | --- | --- | --- |
| [Uncensored AI Roleplay](https://uncensoredllm.ai/models/uncensored-ai-roleplay) uncensored-ai-roleplay | Roleplay | $0.60 | $2.34 |
| [GLM-5.3 Flash Uncensored](https://uncensoredllm.ai/models/glm-5.3-flash-uncensored) glm-5.3-flash-uncensored | Uncensored | $0.1053 | $0.3685 |
| [DeepSeek V4 Pro Uncensored](https://uncensoredllm.ai/models/deepseek-v4-pro-uncensored) deepseek-v4-pro-uncensored | Uncensored | $0.97 | $1.94 |

Prices per million tokens, read from the live catalog on 2026-11-09.

- [Uncensored AI Roleplay](https://uncensoredllm.ai/models/uncensored-ai-roleplay), `uncensored-ai-roleplay`: the roleplay-trained default; the first id we try on a new card.
- [GLM-5.3 Flash Uncensored](https://uncensoredllm.ai/models/glm-5.3-flash-uncensored), `glm-5.3-flash-uncensored`: the low-cost option for long sessions.
- [DeepSeek V4 Pro Uncensored](https://uncensoredllm.ai/models/deepseek-v4-pro-uncensored), `deepseek-v4-pro-uncensored`: for anyone following a "DeepSeek through LoreBary" thread. Give it a response length of 1,000 or more, or a reply that reasons first arrives empty.

The full list is on the [Models page](https://uncensoredllm.ai/models), and the [uncensored LLM leaderboard](https://uncensoredllm.ai/blog/uncensored-llm-leaderboard) ranks it by the daily check at [/proof](https://uncensoredllm.ai/proof) and by price.

## What 100 messages cost

LoreBary adds no fee of its own ("Bring any provider key, we add no fee", on the portal), so the bill is the front-end's prompt at this catalog's prices. A Janitor-sized message is costed at 4,000 tokens in and 300 out, and the same message late in a long chat at 16,000 in. Prices per million tokens come from the live catalog on 2026-11-09.

| Model | One message | 100 messages | 100 messages, 16k context |
| --- | --- | --- | --- |
| `uncensored-ai-roleplay` | $0.0031 | $0.3102 | $1.0302 |
| `glm-5.3-flash-uncensored` | $0.0005 | $0.0532 | $0.1795 |
| `deepseek-v4-pro-uncensored` | $0.0045 | $0.4462 | $1.6102 |

The smallest top-up of $10 buys about 3,223 messages of that size on `uncensored-ai-roleplay` at today's price, and prepaid credit does not expire. The [pricing page](https://uncensoredllm.ai/pricing) has every model and the bonus tiers.

## Check the key before you put LoreBary in front of it

With two services in the chain, a failure needs to be placed before it can be fixed. This request goes to the API directly, in the shape a front-end sends; a 200 here means the key, the URL and the model id are right, and anything that fails afterwards is the hop. Replace the key and keep the rest.

```bash
curl -i https://api.uncensoredllm.ai/v1/chat/completions \
  -H "Authorization: Bearer sk-ullm-live-…" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "uncensored-ai-roleplay",
    "max_tokens": 300,
    "messages": [
      {"role": "system", "content": "You are Oren, a lighthouse keeper on a storm night. Stay in character. Reply in two short sentences."},
      {"role": "user", "content": "There is a light out on the water. Is that a ship?"}
    ]
  }'
```

The answer follows the [chat completions format](https://platform.openai.com/docs/api-reference/chat); `usage` is what the request is billed on, and the `x-uplink-charge` header carries the settled charge. Each request is one line on the console's Usage page whether it came direct or through LoreBary, so the hop never hides the bill.

## What each operator keeps

Through LoreBary a message is seen twice: by LoreBary, whose privacy policy, read 2026-11-02, says request content stays in server memory only for the duration of a single request, and by this API. What this one keeps, as the [privacy page](https://uncensoredllm.ai/privacy) states it: one usage record per request (request id, model, token counts and the charge); SHA-256 fingerprints of each request and response, never the text; and request bodies only when you turn replay on for a key, for 7 days. The key itself is stored as a salted hash and shown once.

A key that sits in a third party's configuration deserves a monthly spend limit, set in the console; the API then answers 402 `monthly_limit_reached` instead of spending on. One key per app keeps a revoke from touching the others.

## Troubleshooting

| What you see | Cause | Fix |
| --- | --- | --- |
| 401, `invalid_api_key` | The key is wrong or revoked, or pasted into the URL. | Paste it again from the console into the key field. |
| 404, `route_not_found` | The endpoint and the appended path do not add up to `/v1/chat/completions`. | Endpoint `https://api.uncensoredllm.ai/v1`, path `/chat/completions`; re-create the proxy. |
| 404, `model_not_found` | The Model field holds an id from another provider's list. | Paste an id from the Models page. |
| Commands or plugins do nothing | A custom proxy relays the chat only. | Expected on this route; ask LoreBary for a permanent route, or go direct. |
| Janitor still answers with its own model | The new proxy settings were not saved or the page was not reloaded. | Press Add and Save, then reload the page. |
| An empty reply | A reasoning model cut inside its reasoning step. | Raise the response length to 1,000 or more. |

```
HTTP 401, the key is wrong
{"error":{"message":"The API key is not valid.","type":"authentication_error","param":null,"code":"invalid_api_key"}}

HTTP 404, the URL does not reach the chat completions route
{"error":{"message":"No such route.","type":"not_found_error","param":null,"code":"route_not_found"}}

HTTP 404, the Model field names an id that is not on the list
{"error":{"message":"No such model: deepseek-chat.","type":"not_found_error","param":null,"code":"model_not_found"}}

HTTP 402, the balance is empty
{"error":{"message":"The prepaid balance does not cover this request. Top up at … to continue.","type":"billing_error","param":null,"code":"insufficient_balance"}}
```

## Questions

### What is LoreBary?

A roleplay platform by Sophia with a proxy layer that sits between a front-end such as Janitor AI, Chub or Wyvern and the model API, adding commands, plugins, lorebooks and summaries on the way through. You bring the model with your own key.

### Does LoreBary charge a fee on my API key?

No. Its Proxy Portal, read 2026-11-02, says it adds no fee; the bill is the front-end’s prompt at the API’s per-token price, on the balance behind the key.

### What proxy URL do I put in Janitor AI when using LoreBary?

The URL LoreBary hands you after Create Proxy, not the API’s own address. For a direct setup without LoreBary it is https://api.uncensoredllm.ai/v1/chat/completions. See the Janitor AI proxy setup.

### Do LoreBary commands and plugins work with a custom proxy?

Not on a custom proxy, according to the older community guide (lorebary-guide.carrd.co, read 2026-11-02, marked outdated on its own page): it limits them to LoreBary’s permanent provider routes. A custom proxy relays the chat and nothing else.

### Can I use an uncensored model on LoreBary?

Yes, through My Custom Proxy with the API as the endpoint and an id such as uncensored-ai-roleplay in the Model field; or directly in the front-end, which is simpler when the model is all you want. See the Models page.

### Is LoreBary safe to put a key into?

LoreBary’s privacy policy, read 2026-11-02, says request content stays in memory for one request. Treat the key like any key handed to a third party: its own key, a monthly limit on it, revoked from the console the day you stop using the route.

Every model above is on the [uncensored model list](https://uncensoredllm.ai/models) with its live price per million tokens. [Per-token pricing](https://uncensoredllm.ai/pricing) sets the whole catalog side by side, and the [family directory](https://uncensoredllm.ai/uncensored) groups the builds by base model.

[Get a key](https://app.uncensoredllm.ai/console) Prepaid from $10, no subscription.

## New models and price changes, once a month.

New uncensored builds with their prices, price moves, and what the daily model check found.