B.AI (@BAI_AGI)

灵活选择，高效赋能 | 订阅与充值方案指南 为满足开发者与 AI 爱好者多样化的使用场景， 提供「按需充值」与「计划订阅」两种灵活模式，让每一份算力精准服务于您的创意与工作。 1️⃣ 随用随充：适合初探与轻量级用户 • 按量计费：用多少充多少，余额灵活使用 • 限时福利：充值即享 1:1 赠送额度，五折畅享全量模型 • 全能准入：支持调用所有主流模型，并解锁完整上下文等专属权限 2️⃣ 订阅计划：适合高频开发者与进阶玩家 • Plan Pro（$200/月）：适用于个人开发者、高频 AI 用户 • Plan Max（$2,000/月）：适用于深度使用者 无论您是偶尔迸发灵感的个人开发者，还是构建生产力的团队， 总有一款方案与您契合。 更多详情见：

已全面接入中国主流支付体系 🇨🇳 微信支付、支付宝与银联现已在 全面可用。这意味着用户可以在熟悉的支付环境中，完成从充值到使用顶级 AI 模型与 Agent 服务的完整流程，让 AI 使用体验更加顺畅、自然与本地化。 🌐 更低门槛，更高效率，让 AI 更轻松走进每个人的日常 →

Graduated from @EASYResidency S3! 🎓 joined by @bankofai_io, is representing the “Agentic Finance” frontier alongside an incredible group of builders. We’re building the infrastructure layer for autonomous AI Agents: Identity, Payments, and Token Fuel for true on chain autonomy. The autonomous economy is just getting started. 🚀

Claude Series Model Maintenance Notice To ensure service stability and billing accuracy, the platform will conduct maintenance and verification for the Claude series model service pipeline. During the maintenance period, Claude series models may be temporarily unavailable or experience request failures. Other model services will remain unaffected, and users may switch to other available models such as GPT, DeepSeek, and Gemini. We will restore Claude series model services as soon as the maintenance is completed. Thank you for your understanding and support. Claude 系列模型服务维护通知 为保障模型调用稳定性及计费准确性，平台将对 Claude 系列模型相关服务链路进行维护校验。 维护期间，Claude 系列模型可能会暂时不可用，或出现调用失败等情况。其他模型服务不受影响，用户可正常切换使用 GPT、DeepSeek、Gemini 等其他可用模型。 维护完成后，我们将第一时间恢复 Claude 系列模型服务。感谢大家的理解与支持！

From Infrastructure to Interface: Closes the Loop. In response to community demand, we have officially synchronized our Web Chat with our API ecosystem. The four frontier models—including GPT-5.5-Instant, DeepSeek-V3.2, MiniMax-M2.7, and GLM-5.1—are now fully accessible to all web users. We have bridged the gap between developer-grade integration and consumer-facing interaction. Whether via API or Web, you can now experience the same production-grade reliability and reasoning consistency. Compute without limits. Innovate without boundaries. Start now: 从底层基建到场景应用： 正式打通全链路模型闭环。🌐 应社区用户的热切期待， 现已完成 API 与 Web Chat 的双端能力同步。此前在 API 侧首发的 GPT-5.5-Instant、DeepSeek-V3.2、MiniMax-M2.7 以及 GLM-5.1 四大顶尖模型，现已在网页端全量上线。 无论你是追求高自由度的开发者，还是深耕生产力的专业用户，现在都能在 获得一致的生产级可靠性与逻辑精度。 算力不设限，灵感无边界。 立即体验：

📢 关于 历史 Credits 消耗记录优化与补偿说明 随着 Credits 计费体系与展示维度的持续升级，我们近期对历史消耗记录进行了全面复核，以进一步提升透明度与一致性。在复核过程中，我们发现：在部分早期请求中，前端展示的消耗明细未完整呈现全部成本构成，主要是部分底层缓存相关费用未在页面中展开展示。 因此，用户在历史记录中通常只能看到 input / output token，从而在理解整体 Credits 消耗时出现一定偏差。本次问题已完成统一修正与历史数据重新对齐，并基于更谨慎、更用户友好的规则，对受影响账户进行 Credits 补偿。 📅 覆盖时间范围： 2026 年 5 月 1 日 00:00 至 5 月 3 日 03:00（SGT） 💳 补偿说明： Credits 将自动发放至受影响账户，无需任何申请操作，请留意账户余额更新 ⏳ 有效期说明：本次补偿 Credits 有效期为 30 天，自到账之日起计算 与此同时，我们也将持续升级 Credits 展示体系，使整体消耗结构在 input、output 及底层成本项维度上更加清晰统一，提升长期透明度体验。 感谢大家的理解与持续反馈！

📢 100B Token Subsidy — Limited-Time Drop 🔹 1:1 deposit match: every top-up is instantly doubled in Credits 🔹 Up to $100 bonus credits per user (cumulative, until cap is reached) 🔹 $10,000 daily subsidy pool: resets every day at 00:00 SGT, first come first served This is a full subsidy mode for AI execution at scale. Deposit now and run top-tier models like GPT-5.5, Claude Opus 4.7, and DeepSeek V4 at up to half the cost, powering real AI agents and production workflows with significantly lower spend. 🔥 New users get 500,000 FREE Credits instantly. 🏠 Build more. Run more. Pay less. Scale intelligence →

📢 “白”亿补贴｜百亿 Token 限时放送中！ 🔹 充值即返：每笔充值自动获赠等额 Credits 🔹 最高赠送 $100：不限充值次数，送满即止 🔹 每日 $10,000 补贴池：每日凌晨重置（SGT），先到先得 海量 AI 能力现已全面进入“5折补贴模式”。现在充值，用更低成本，畅用DeepSeek V4、 GPT-5.5、Claude Opus 4.7等顶尖模型，让你的 AI Agent 真正跑起来！ 🔥 新用户可领 500,000 免费 Credits! 🔗 来 “白” 用 AI，白拿、白跑、白体验 →

📢 API Model Arsenal Expanded: GPT-5.5 Instant & Three Other Frontier Models Are Now Live Committed to building the most efficient underlying AI compute network for developers, has successfully integrated and deployed four of the latest models to our API endpoints: DeepSeek-v3.2, MiniMax-M2.7, GLM-5.1, and GPT-5.5-Instant. 🔹 Rapid Underlying Integration: Following OpenAI's recent official release of GPT-5.5 Instant, the team completed the full adaptation of underlying routing and interfaces within 48 hours, ensuring developers have zero-delay access to next-generation computing capabilities. 🔹 Leap in Production-Grade Reliability: As the latest architectural model, GPT-5.5 Instant delivers enhanced logical clarity and output conciseness. In highly rigorous workflows such as medicine, law, and finance, its hallucination rate has decreased by 52.5% compared to its predecessor (GPT-5.3 Instant), significantly elevating factual accuracy for complex tasks. The aforementioned models are now fully accessible via API requests. The Web Chat interface will also follow suit with these updates—stay tuned. Review official documentation and connect to the latest APIs: 🔗 📢 API 模型库极速扩容：GPT-5.5 Instant 等四大模型现已就位 致力于为开发者构建最高效的底层 AI 算力网络， 现已在 API 侧完成四款最新模型的集成与上线：DeepSeek-v3.2、MiniMax-M2.7、GLM-5.1 与 GPT-5.5-Instant。 🔹 极速底层集成： OpenAI 于日前正式发布 GPT-5.5 Instant， 研发团队在 48 小时内即完成了底层路由与接口的全面适配，确保平台开发者实现新一代算力的零时差调用。 🔹 生产级可靠性跃升： 作为最新架构模型，GPT-5.5 Instant 具备更强的逻辑清晰度与输出精简度。在医疗、法律、金融等高严谨度业务流中，其幻觉率较前代（GPT-5.3 Instant）降低 52.5%，大幅提升了复杂任务的事实准确性。 上述模型现已通过 API 开放全量请求访问。Web 对话端也将跟进上线，敬请期待。 查阅官方文档并接入最新 API： 🔗

📢 1.7M Users on 🔥 A huge thank you to the community. We are accelerating the path to AGI by building the ultimate infrastructure for privacy, smart routing, and the Agent economy. Powered by: ▪️ 500,000 Free Credits for newcomers ▪️ Limited-time 1:1 Match & 50% cost reduction ▪️ Zero Markup Across All Models ▪️ Flexible top-ups & subscriptions We aren't just giving access to AI; we are building the execution layer where autonomous intelligence scales. The march to 2,000,000 starts now. 🚀 👉

📢 体验升级：付费用户专属「全量上下文」权限已解锁！ 告别字数限制，释放大模型的真正阅读极限。 即刻起， 面向所有充值用户，正式开放 API 与 网页Chat 端的全量上下文能力： ▪️打破人为内存上限， 彻底解除基础字符限制。 ▪️专为重度任务而生， 轻松实现超长文档分析、万行级代码库审查与海量数据提炼，告别信息断层与遗忘。 不再被长度束缚，将最硬核的分析任务交给

📢 Model Update: DeepSeek-V4 Series Now Available on To continuously push the boundaries of developer capabilities, has officially integrated the latest DeepSeek flagship series. 🧠 DeepSeek-V4-Pro: Engineered for extreme reasoning, complex coding, and heavy logic workloads. ⚡️ DeepSeek-V4-Flash: Lightning-fast response with ultra-low latency. Official documentation is updated and ready: 👉Start building with the latest intelligence: 📢 模型库更新：DeepSeek-V4 系列现已在 上线 为持续拓宽开发者的算力边界， 现已正式上线 DeepSeek 最新旗舰系列。无论您是在处理复杂逻辑，还是追求极致的并发速度，总有一款适合您的 Agent： 🧠 DeepSeek-V4-Pro： 专为极度复杂的逻辑推理与代码生成打造。 ⚡️ DeepSeek-V4-Flash： 极致的响应速度与超低延迟。 📚 开发者接入文档已同步更新： 👉即刻调用最新算力，探索极致效能：

📢 计费引擎全面升级：Prompt Caching（上下文缓存）精准计费现已上线 全面兑现「全线模型官方零差价」的承诺，我们的底层计费系统在今日凌晨完成了重要升级，全面适配最新大模型的缓存计费机制： 🔹 精准的缓存折扣： 系统现已支持动态识别上下文缓存命中状态。当您的请求命中缓存时，将自动享受等同于官方的调用成本（最高可节省 90% ）。 🔹 计费核算更准确： 底层费率统计逻辑已全面修复与更新。现在，您在 Usage 页面看到的每一笔 Credit 消耗，都已精准包含缓存命中后的抵扣，真正的零差价，每一分算力消耗都清晰透明。 无需修改代码，立即享受更低成本、更低延迟的智能体交互体验。 探索最新模型： 🔗

📢 Billing Engine Upgraded: Accurate Prompt Caching is Live Delivering on our promise of "Zero Markup" across all models, we deployed a major upgrade to our billing infrastructure today to perfectly adapt to the caching mechanisms of the latest frontier models: 🔹 Accurate Cache Discounts: Our system now dynamically detects context cache hits. When your request hits the official cache, you will automatically enjoy the exact same discount as the official pricing (saving up to 90% on API costs). 🔹 Precise Billing Calculation: The underlying rate statistics logic has been completely updated. Every credit deduction you see on the Usage page now accurately reflects the discounted cache rate. True zero markup, with every token calculated transparently. Enjoy lower costs and lower latency for your AI Agents. 🔗