Relay365 FAQ: Claude Code pricing on one API relay desk

What the Relay365 duty desk actually runs

Relay365 is an independent third-party relay run like a round-the-clock duty desk. Every lane sits on one table there: Claude, GPT, Gemini, DeepSeek and the rest are reached through the same account, the same wallet and the same key management. For a careful buyer, that is the thing to settle early, because a relay that scatters you across several vendors turns a single budget decision into a small operations project. Relay365 keeps the desk in one place.

The commercial terms are deliberately plain. Relay365 charges per 1 million tokens in approximate USD, while the account itself is kept in CNY. There is no monthly fee: the wallet is prepaid, and every call is settled on the token counts the upstream reports. A new account starts with $0.25 in signup credit, so you can open a Relay365 account, look at the pricing page and wire a small test without a long commitment.

The other question a careful buyer asks early is what happens when something goes wrong. A call that fails upstream with a 502 or a 503 is not billed. A disconnected stream is billed for what was actually used. Those two rules are short, but they are the ones that decide whether a relay feels like a desk watching the line or like an invoice that arrives later.

How the four token types are priced

Relay365 prices input, output, cache write and cache read separately, and every request can be traced in the usage log. A careful buyer should not accept a single blended number, because a workload that reads cached context behaves very differently from one that writes it. Here is how the table reads on a few lanes.

• Claude Fable 5, wire name claude-fable-5, is $3.79 input, $18.94 output, $4.66 cache write and $0.2917 cache read per 1 million tokens. Official list price is input $10, output $50 and cache read $1, so the Relay365 output rate is about 38% of official.
• GPT-6 Astra, wire name gpt-6-astra, is $0.7444 input, $3.72 output and $0.075 cache read. Official list price is input $10, output $50 and cache read $1, so the Relay365 output rate is about 7% of official.
• DeepSeek V4 Pro, wire name deepseek-v4-pro, is $1.45 input, $4.35 output and $0.0486 cache read. No official price is quoted for this family, so Relay365 makes no comparison here; the numbers on the pricing page stand on their own.
• Gemini 2.5 Flash, wire name gemini-2.5-flash, is $0.2389 input, $1.91 output, $0.2389 cache write and $0.0236 cache read. No official price is quoted for this family either, so again the table speaks for itself.

To make the arithmetic concrete: for example, if you send 1 million input tokens and 1 million output tokens a day on Claude Fable 5, that is $3.79 plus $18.94, or $22.73 a day. The same day on GPT-6 Astra is $0.7444 plus $3.72, or $4.4644. On DeepSeek V4 Pro it is $1.45 plus $4.35, or $5.80. These are examples, not quotes; your own bill follows the token counts the upstream reports.

Which clients each lane will accept

Relay365 is strict here, and the desk would rather say so plainly than have a buyer discover it mid-integration. Claude models work only from the official Claude Code and Claude Desktop clients, and Claude Code 2.1.285 or newer is required. They do not work from Cursor, Cherry Studio, Cline, OpenCode, any SDK or cURL; those clients pair only with GPT, DeepSeek, Kimi, GLM, Qwen, Grok, MiniMax and Hunyuan lanes.

Gemini has its own rule. It runs on its native protocol only, through Gemini CLI, and it is not OpenAI-compatible. A single reply goes to :generateContent, and a stream goes to :streamGenerateContent?alt=sse. It cannot go through /v1/chat/completions. If your current setup already speaks the native Gemini protocol, Relay365 will feel familiar; if it does not, the desk can point you at the lanes that fit your client.

One more detail worth asking about: a key can be bound to a lane, and keys can carry tags. That keeps a production key separate from a scratch key and makes the Relay365 table easier to read at a glance.

Duty passes, the wallet and the money questions

Relay365 offers optional duty passes for Claude and GPT only; every other model always uses the wallet. A pass comes with a dedicated key that never touches the wallet, so pass traffic and wallet traffic stay cleanly separated in the ledger.

• The Claude pass runs 14 days and costs $50.97. It draws up to $700 of quota at the duty-pass rate x2.5, which is at most about $280 of usage at list prices, and less on models whose pass rate sits above list. The cap is $380 of quota per 7 days.
• The GPT pass runs 30 days and costs $81.53. It draws up to $3000 of quota at the duty-pass rate x2.5, at most about $1200 of usage at list prices, and less on models whose pass rate sits above list. The cap is $900 of quota per 7 days.

The referral bounty is worth understanding before you invite anyone. It pays 18% at level 1 and 6% at level 2 of what referred members pay. It can be withdrawn from $11.11 or moved into the wallet.

Refunds are clear too. Within 14 days of a top-up, unused balance can be refunded minus what was consumed and the payment channel's fees. Bounty and promotional credit are not refundable, so treat those as spendable rather than cash-like. The Relay365 desk will walk through any of these questions in writing before you top up.

Limits, logs and reaching the desk

Know the limits before you build a burst pattern around them. An HTTP 429 appears when a single key makes more than 120 calls in 60 seconds, or when one account has more than 8 requests in flight. Those are account-level facts you can design around, and the desk will point them out before you hit them.

The dashboard is where the round-the-clock part becomes useful. Relay365 keeps a usage log of the latest 200 requests showing the time, model, input, output, cache recall, charge and result of each call. It does not show which key made the call, so plan on separating lanes by key rather than by log line. Alongside it, a ledger of wallet, bounty and pass accounts keeps the balances in one view, quick-wiring snippets shorten setup, and the interface is available in six languages.

And to reach the desk, you do not need to hunt for an address. Relay365 has in-site messages to the desk, available once you log in. Send the question you actually have — a rate, a client rule, a refund — and the answer stays attached to the account you are working in.

If the desk has answered your questions, create a Relay365 account, use the $0.25 signup credit, and send your first request through the lane you plan to use.