Bybit V5 API 限频响应头：重试预算清单

NexuAlpha 直接答案：Bybit V5 API throttling is not a single number. The current official rate-limit page separates HTTP IP limits, WebSocket IP connection limits and endpoint-level per-UID API limits, and it exposes response headers that describe the current endpoint budget. A production client should budget retries from those headers instead of reconnecting blindly after every rejection.

决策结果：Treat a Bybit API integration as ready only after documenting IP budget, UID endpoint budget, WebSocket connection budget, error handling for retCode 10006, a 403 cooldown rule and a reconciliation path after any throttled order action.

证据快照。本文在 2026-09-21T08:43:24.237Z 读取 Bybit API Documentation。本地队列、旧草稿、搜索摘要、社交转述、价格图或 AI 摘要都不被当作事实证明。它们最多提示问题，不能替代本文公开来源列表中的来源映射。

核验事实。主张 c1：Bybit documents a default HTTP IP limit of 600 requests within a five-second window per IP for traffic to api.bybit.com and local API hostnames. 主张 c2：Bybit says an access-too-frequent 403 response means the IP exceeded allowed request frequency and the application should terminate HTTP sessions and wait at least ten minutes. 主张 c3：Bybit documents WebSocket IP limits, including not establishing more than 500 connections within a five-minute window. 主张 c4：Bybit says endpoint API limits are based on a rolling one-second window per UID and exposes X-Bapi-Limit, X-Bapi-Limit-Status and X-Bapi-Limit-Reset-Timestamp headers.这些事实刻意保持窄口径，不证明个人资格、未来可用性、执行质量、超出原文的法律状态、税务处理、账户批准、客服结果、赎回成功、价格方向、流动性深度或平台偿付能力。

围绕 Bybit V5 API rate limit headers retry budget checklist 的决策纪律，不是寻找绝对赢家，而是建立来源绑定的评分卡：哪个官方来源回答了当前问题，哪个控制项仍为空白，哪个空白会改变用户风险，哪个测试可逆，哪个条件会重开决策。诚实的空白比编造优势更安全。

FAQ。短答案是什么？Bybit V5 API throttling is not a single number. 首先核验什么？核验官方来源和准确控制表面。不能推断什么？不能推断资格、未来支持、法律处理、税务处理、恢复、执行质量、流动性、赎回、消费者保护或账户批准。何时停止？当来源、对象、用户状态、权限、路径、发行人、司法辖区或故障处理不清楚时停止。

公开文档不是私人账户保证。 证明合理决策不应要求暴露秘密。

可见的空白能阻止读者用乐观填补缺口。

公开文档不是私人账户保证。 证明合理决策不应要求暴露秘密。

可见的空白能阻止读者用乐观填补缺口。