BestRelay pay-as-you-go: GPT-6.1 Sol at $0.2083 input rate

BestRelay usage records show every call with its model and exact charge

You open the BestRelay dashboard at the start of a workday and the usage records are already there, each row a finished request. BestRelay is an independent API relay that forwards every request to the model provider under the exact model name you wrote, so the model named is the model that answers and nothing is quietly swapped for a cheaper tier. The line-up is curated: 20 public models, each listed under its own name with four prices for input, output, cache write and cache read where applicable. When a provider retires a model id, BestRelay delists it. Billing is prepaid pay-as-you-go balance with 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. Nothing is held in advance. Failed calls that the provider rejects are recorded at zero, and when the provider cannot be reached nothing is charged. Every call appears in the usage records with its model, tokens and charge, giving you a complete ledger for the day.

For example, imagine you spend a morning with Sonnet 5.5. The usage records show 500,000 input tokens and 100,000 output tokens for a single day, and the charge is $1.40 on BestRelay. If you repeat that pattern for 22 such days the total is $30.71. When 350,000 of the input tokens are read from cache the day costs $0.94. At the official list price the same day would be $2.00, the 22 days $44.00 and the cached day $1.33; the difference on the day is $0.60 and on the 22 days $13.29. These figures come directly from the worked examples on the pricing sheet and illustrate how the records let you verify every number.

BestRelay pricing appears in the records as concrete per-million rates

The usage records do not hide the arithmetic. Each model in the curated line-up carries its own rate card. Sonnet 5.5 lists input at $1.39, output at $7.01, cache write at $1.74 and cache read at $0.0944 per 1,000,000 tokens. Opus 5.5 shows input $1.74, output $8.13, cache write $2.01 and cache read $0.0778. Haiku 4.5 is input $0.4167, output $2.01, cache write $0.4861 and cache read $0.0333. GPT-6.1 Sol offers input $0.2083, output $1.11 and cache read $0.0431. DeepSeek V4 Pro lists input $1.53, output $4.58 and cache read $0.0514. Gemini 3.1 Pro carries input $1.74, output $9.93, cache write $1.74 and cache read $0.1389. Kimi K3 shows input $1.39, output $7.01 and cache read $0.1389. Grok 4.7 is input $0.625, output $1.94 and cache read $0.1389. GLM-5.3 mirrors that at input $0.625, output $1.94 and cache read $0.1389. When you scan the day's rows you see each call charged at exactly these rates, multiplied by the tokens the provider reported. There are no hidden mark-ups and no subscription fees to amortise.

For example, a day with GPT-6.1 Sol at 500,000 input and 100,000 output tokens costs $0.22 on BestRelay, and 22 such days cost $4.74. With 350,000 of the input read from cache the day costs $0.16. At the official list price the same day is $2.00, the 22 days $44.00 and the cached day $1.33; the difference on the day is $1.78 and on the 22 days $39.26. The free credit alone covers about 1.6 million GPT-6.1 Sol input tokens. These worked examples are quoted from the pricing sheet so you can match them against your own records line by line.

BestRelay keys and limits keep the records clean and attributable

An account can hold several keys, each created under API keys in the dashboard, and each key can be voided at any time. Every call appears in the usage records tied to the key that made it, so you can separate projects or environments without mixing their charges. Limits are 120 requests a minute on one key and 8 calls in flight on one account. These caps are enforced by the relay and are visible in the records as rejected requests, which are recorded at zero charge. Connecting a client needs only two settings: the address and a key. Details are in the docs. The records also respect privacy: the user-written text of requests is kept for 30 days to review abuse and settle billing questions; system prompts and model replies are not stored.

Client compatibility is explicit in the records because the model name is fixed. 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. When you filter the records by model you immediately see which client produced each row, and there is no ambiguity about whether a Claude call came from an unsupported tool.

BestRelay top-ups and refunds appear as balance events beside the call logs

The balance timeline sits next to the usage records. Top-ups are processed through Stripe checkout in fixed amounts: $1.39, $6.94, $13.89, $27.78, $69.44 and $138.89. Each new account opens with $0.35 of credit. 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. Referrals earn 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. As an example, on an invited member's top-up of $13.89 the two commissions are $2.50 and $0.5556. All of these events are dated and amount-stamped in the same view that shows your calls, so the financial picture is complete without exporting spreadsheets.

BestRelay records let you close the day with a verified total

At the end of the sample day you scroll through the usage records. Each row states the model, the input tokens, the output tokens, the cache read tokens where applicable, and the charge calculated from the published rates. The sum at the bottom matches the balance deduction to the cent. There are no mystery fees, no monthly minimums, no subscription charges and no advance holds. The prepaid balance simply decreases by the exact amount of each completed call. Failed provider rejections show a zero charge. Network unreachable events show no charge at all. The model name in each row is the model you requested, and because BestRelay forwards under that name the answer comes from that model. The curated line-up of 20 models, each with its own rate card, means you always know which price applies. The keys you created keep projects separate. The 30-day retention of request text gives you a window to audit any dispute. The referral commissions and top-up history sit in the same ledger. You close the dashboard with a ledger that balances, and the next morning the same transparency awaits. Try BestRelay and see the records for yourself.