BestRelay API relay for Claude Fable 5.1 pay-as-you-go pricing

BestRelay sends your request to the model whose name you wrote

At the center of BestRelay sits one rule that shapes everything else: the model you name is the model that answers. BestRelay is an independent API relay, and it forwards each request to the model provider under the model name the caller wrote. Nothing is quietly swapped for a cheaper tier, and no silent substitution stands between the name in your request and the model that returns the reply. That single fact is why the line-up matters, and BestRelay keeps the line-up curated: 20 public models, each listed under its own name, each with four prices for input, output, cache write and cache read. When a provider retires a model id, BestRelay delists it, so the names on the rate card are the names that can actually answer.

This is a small, careful publication of a service, and it is proud of that care. BestRelay is a relay and nothing more elaborate than that: it carries traffic to providers and reports what the providers report. What it offers instead of ornament is exactness. The model named is the model that answers, the price shown is the price charged, and the record kept is the record you can read. For a caller who has ever wondered whether a flagship name really produced the output in front of them, that exactness is the product, and it is the reason the rest of this article reads the rate card line by line rather than describing it in adjectives.

The rate card prices Claude Fable 5.1 at $4.10 input and $20.42 output per million tokens

Read line by line, the row for Claude Fable 5.1 on BestRelay is short and exact. Input costs $4.10 per 1,000,000 tokens, output costs $20.42, cache write costs $5.00, and cache read costs $0.2083. The official list price for the same model is input $10, output $50, cache read $0.25, so the rate card on BestRelay sits below the list price on every line the sheet gives. The line-up around it is equally explicit: Claude Fable 5 is $4.17 input and $20.97 output, Claude Opus 5.5 is $1.74 input and $8.13 output, and Claude Sonnet 5.5 is $1.39 input and $7.01 output, all per 1,000,000 tokens, each with its own cache write and cache read.

For example, a day of 500,000 input and 100,000 output tokens on Claude Fable 5.1 costs $4.09 on BestRelay, and 22 such days cost $89.99. With 350,000 of the input read from cache, that day costs $2.73. At the official list price the same day is $10.00, the 22 days are $220.00, and the cached day is $6.59; the difference on the day is $5.91 and on the 22 days is $130.01. For example, the same exercise on Claude Sonnet 5.5 gives $1.40 for the day, $30.71 for the 22 days and $0.94 for the cached day, against $2.00, $44.00 and $1.33 at the official list price, with differences of $0.60 and $13.29. These are examples, quoted as examples, of how the arithmetic reads when a rate card meets a working day.

BestRelay bills once, after the call ends, with no monthly fee

Billing on BestRelay is prepaid pay-as-you-go, and the shape of it is deliberately plain. There is no monthly fee, no subscription and no minimum spend. Each call is charged once, after it ends, from the tokens the provider reports, at the rate card prices, and nothing is held in advance. Requests the provider rejects are recorded at zero, and when the provider cannot be reached nothing is charged, so a failed call never becomes a paid call. Each new account opens with $0.35 of credit, which alone covers about 1.6 million GPT-6.1 Sol input tokens, a modest enough beginning to test the plumbing without ceremony.

Top-ups go through Stripe checkout in fixed amounts: $1.39, $6.94, $13.89, $27.78, $69.44 and $138.89. Unused credit can be refunded on request within 13 days of the top-up, less what was consumed and any payment-processor fees; gift credit and commission cannot be refunded. BestRelay also runs a referral program: 18% of the top-ups of people who joined through your link, and 4% of the top-ups of the people they invited. Commission can be moved into the balance, or withdrawn to Alipay from $20.83. For example, on an invited member's top-up of $13.89, the two commissions are $2.50 and $0.5556. The point of all this bookkeeping is narrow: you can add credit, spend it, and see exactly where it went.

Claude models on BestRelay are served only to Claude Code and Claude Desktop

The client rule on BestRelay is strict, and it is worth reading before you connect. Claude models are served only to the official Claude Code and Claude Desktop clients, version 2.1.285 or newer. They do not work in Cursor, Cherry Studio, Cline, OpenCode, any SDK or cURL. Those tools, like Codex, pair only with GPT, DeepSeek, Kimi, GLM, Qwen and Grok. Gemini models use only their native protocol, through Gemini CLI or direct generateContent calls, and are not OpenAI-compatible. The pairing is therefore fixed in advance: name a Claude model and use Claude Code or Claude Desktop; name GPT-5.6 Sol, DeepSeek V4 Pro, Kimi K3, GLM-5.3, Qwen3.8 Max or Grok 4.6 in another tool and that is the model that answers.

Connecting needs two settings: the address and a key created under API keys in the dashboard. The details live in the docs, and BestRelay keeps them there rather than printing them in an article. Once those two settings are in place, the request goes out under the model name you wrote, and the answer comes back from that same model. Nothing about the arrangement asks you to accept a substituted tier in exchange for a lower number on the rate card, which is the trade-off most relays make and the one BestRelay declines to make.

BestRelay records every call with its model, tokens and charge

On the operations side, BestRelay keeps the limits and the records visible. One key is limited to 120 requests a minute, and one account to 8 calls in flight. An account can hold several keys, each key can be voided at any time, and every call appears in the usage records with its model, tokens and charge, so the bill can always be checked line by line against the rate card. Privacy is stated with the same precision: the user-written text of requests is kept for 30 days to review abuse and settle billing questions, while system prompts and model replies are not stored.

Put together, these pieces make the case for trying BestRelay. The model you name is the model that answers, the price you read is the price you pay, a failed call costs nothing, and the record of every call is there to read afterward. Open an account, take the $0.35 of credit, name Claude Fable 5.1 in Claude Code or Claude Desktop, and let the rate card speak for itself on a real request.