Claude as a Janitor AI Proxy: What It Costs per Message, the Two Routes, and Anthropic's Own Caveat

Claude has no free route on Janitor AI at all. Anthropic sells by the token, OpenRouter lists no :free Claude, and a "free Claude proxy" shared in a Discord runs on somebody else's key. Chub's documentation warns that such proxies may log what you send and use stolen keys. So this guide does something the free-backend guides do not: it starts with the bill. At listed prices Claude is cheaper per message than most people assume, and the arithmetic is the first thing to see.

Two routes then follow, with exact values. And a caveat from Anthropic itself about the more direct of the two, quoted rather than paraphrased, because it decides which route to pick.

What a message costs, at listed prices

Every request Janitor sends carries the whole context: the character definition, your persona, the chat memory, the advanced prompt and the recent history. You pay for all of it as input tokens, every message, plus the reply as output tokens. On 28 September 2026 OpenRouter's public model list priced the current Claude models like this:

Sonnet 5.5 joined the list that evening, at the price of Sonnet 5. Opus 5.5 is the newer Opus and the cheaper one. The list still carried anthropic/claude-sonnet-5 at $2 and $10, anthropic/claude-opus-5 at $5 and $25, and anthropic/claude-sonnet-4.6 at $3 and $15.

The arithmetic for a heavy roleplay request of 20,000 input tokens and a 300-token reply: on Sonnet 5.5, 20,000 tokens at $2 per million is $0.04, and 300 output tokens at $10 per million is $0.003, so about four cents a message. The same message costs about two cents on Haiku 4.5, nine on Opus 5.5 and 22 on Fable 5.1. A hundred messages an evening on Sonnet 5.5 is roughly $4. The same evening on a :free OpenRouter model is $0, and on DeepSeek's own API it is 32 to 64 cents at its listed prices. Claude is the expensive seat in the room, and whether its writing is worth it is a taste question this page does not settle.

Context length is the lever. Trim the chat memory and the advanced prompt and a message that cost four cents costs two.

The two routes to Claude in Janitor AI

Both end with Claude writing your replies. They differ in setup and in which account holds your credit.

Janitor also runs janitor+ Router, which "lets you use third-party AI models directly inside JanitorAI, without setting up an external proxy or pasting in API keys from another service." It is for janitor+ subscribers on the web and is billed per token. Its help article names no models, so look for Claude in the model list in your settings.

Route one: OpenRouter

The route built for this. One OpenRouter account, a balance loaded, and the Claude models sit behind the same URL as everything else.

OpenRouter deducts the listed price from your balance per request. Its limits page says a negative credit balance can bring 402 errors, even on free models, until credits put the balance above zero. There is no daily cap on paid models the way there is on :free ones. If you already use OpenRouter for free models, this is a model-id change and a balance, nothing else; the general setup is in our OpenRouter guide.

Route two: Anthropic's own compatibility layer

Anthropic's API is not natively OpenAI-shaped, which is why Janitor could not talk to it for a long time. It now can, through a compatibility layer that Anthropic documents: point an OpenAI client at the base URL https://api.anthropic.com/v1/, use a Claude API key, and use Claude model names. In Janitor's terms:

Read Anthropic's own note before choosing this route. The documentation says the compatibility layer is primarily intended to test and compare model capabilities, and is not considered a long-term or production-ready solution for most use cases, while adding that it is intended to remain fully functional and not have breaking changes. The documented differences matter for roleplay: all system and developer messages are concatenated into a single system message at the start, temperature above 1 is capped at 1, several OpenAI fields are silently ignored, and rate limits follow Anthropic's standard limits for its messages endpoint, which are per organisation. For one person's roleplay it works; for anything you want to keep working without surprises, route one is the supported path.

What a proxy of any kind does not change

The usage policy. Anthropic's usage policy governs what Claude will write, and it applies to the model wherever the request came from. Janitor is known for uncensored roleplay; Claude is not an uncensored model, and a refused scene is Claude declining, not a broken setting. Changing the route, the URL or your IP address does not change the policy.

The rate limit. Through OpenRouter, the limit is your OpenRouter account's; through Anthropic, it is your organisation's tier. Both follow the key. Our 429 page covers the general rule.

The bill. Prices are per token and per model, and no route makes Claude free. A "free Claude proxy" runs on somebody else's key. Chub's documentation puts it directly: reverse proxies "are created by unauthorized third parties who gain access to your data and IP address. They may enable logging, leak your information, and use stolen API keys which you have no control over." Do not.

Where a network proxy fits

A network proxy changes the address your traffic leaves from, and nothing on this page needs one. Anthropic lists where it offers Claude.ai and commercial API access. On 28 September 2026 both lists named the same 185 countries, regions and territories. Mainland China, Hong Kong, Macau, Russia and Iran were not among them. Janitor AI has one rule that follows the address. Users connecting from Australia, Brazil or the UK must pass a one-time age check, and Janitor says "Detection is IP-based only." Before that check, it blocked UK addresses from 24 July 2025 until 15 June 2026. If Claude is not offered in your country, or your school or office network filters api.anthropic.com, openrouter.ai or janitorai.com, we give no steps around that.

Where our lines do fit is a network that needs a known exit, such as a company network. For that, a static residential or ISP address in the country where you are costs from $0.44 per GB on a usage meter. Our proxies for Claude AI post covers the Claude case: a fixed address an Enterprise admin can put on the Claude IP allowlist. For cost, limits, refusals and keys, our product does nothing. The honest recommendation is to load $5 on OpenRouter, pick anthropic/claude-sonnet-5.5, and watch the balance.

Sources

• OpenRouter, public models list: Claude model ids and per-token prices, read 28 September 2026; API rate limits: 402 on a negative balance.
• Anthropic, OpenAI SDK compatibility: the base URL, the testing-and-comparison caveat, system-message hoisting, the temperature cap, ignored fields and rate limits; models overview: the current model names; rate limits: limits set per organization.
• Anthropic, supported countries, read 28 September 2026: the two lists of 185; usage policy.
• Chub documentation, API connections: the warning about third-party reverse proxies.
• Janitor AI Help Center, The Absolute Beginner's Guide to Using a Proxy with Janitor, Part One, read in its Web Archive copy of 31 July 2026 (Janitor removed the guide by 28 September 2026): the field names; janitor+ Router: Janitor's own per-token route; Age Verification for Australia, Brazil, and the UK: detection by IP address.
• Janitor AI, Tough news for our UK users (July 2025) and An update for users in Brazil, Australia, and the UK (15 June 2026): the UK block and the switch to age verification.