乐天图书馆

[Model Maintenance] The gemini series models are under temporary maintenance, expected to be restored before midnight tonight. Please use other models in the meantime. Thank you for your understanding.

【New Models】Many vendors have been releasing new models recently, and we can't fall behind. DeepSeek-V4-Pro-0813, glm-5.3, and grok-4.6 are now available. Welcome to try them out! Additionally, DeepSeek will officially raise prices by 2-5 times next Monday. We appreciate your long-term support and will do our best to keep prices unchanged!

[New Model] Today we launched the cost-effective gemini-3.5-flash-lite from the Gemini series, and also fixed the recent network fluctuations of gemini-3.6-flash. It now performs well and fast, and supports web search! The price has also been lowered—welcome to use it!

[New Model] Musk's latest reasoning model grok-4.5 is now available. Welcome to try it. Reasoning models take a bit longer to respond; if speed is important, we recommend sticking with the older models.

[New Model] DeepSeek's latest model DeepSeek-V4-Flash-0731 is now available, with significantly improved overall capabilities. Welcome to try it!

[Price Drop] Thanks to ample supplier channels, the strongest domestic model kimi-k3 has been temporarily discounted to 20% of the official price. Same performance at a lower cost—its results surpass claude-4.8-opus, yet it costs only 1/3 as much. Hurry up and enjoy!

[Recharge Restored] Both Alipay and WeChat direct recharge channels on the website have been restored. You can choose either. Still, we recommend contacting customer service QQ: 746999675 for recharges—returning customers get extra bonuses.

[New Model] The wildly popular top domestic model kimi-k3 is now available. Its front-end performance surpasses the Claude series and ranks first in the world. Welcome to try it!

[God-tier Model] Because its performance is absolutely explosive and the US prohibits other countries from using it, the Claude-series claude-5-fable model is now online with a bang! Not sure how long it will last, so use it while you can! Also, the new top-tier GPT series models gpt-5.6-luna and gpt-5.6-sol are live. Among them, gpt-5.6-sol approaches claude-5-fable's performance at a much lower price.

[Model Maintenance] The Gemini series is under official risk control today. Maintenance will be completed by tomorrow at the latest. Thanks for your understanding.

[New Model] GPT's latest model gpt-5.6-terra is now available, with performance that fully surpasses Claude. Welcome to try it!

[Model Fluctuation] We've been updating domestic model channels over the past few days, so occasional fluctuations may occur. Everything should gradually return to normal within the next day or two. We appreciate your patience.

[New Model] Added kimi-k2.7-code and minimax-m3. The Claude replacement works quite well for coding—welcome to try it. Also, the check-in system has been badly damaged, and it will probably take a long time to fix the bugs. We're thinking about how to continue welfare events, so please stay tuned.

[Check-in Maintenance] The check-in system https://api.lotte-library.top/checkin has been attacked by some freeloaders, preventing everyone from checking in and claiming rewards. It's currently under maintenance (expected to be done within three days). Please be patient.

[Fix Update] We're continuing to launch the gpt reverse-proxy special-priced group. Compared to the previous general-purpose models, these models are better suited for everyday text tasks like conversation, search, and translation. For tool-based needs like Codex and ClaudeCode programming, we still recommend the previous general version. The new version is over 60% cheaper than the general version, offering great value. The new model IDs are: gpt-5-1, gpt-5-2, gpt-5-3, gpt-5-3-mini, gpt-5-5. To distinguish: the new versions use a hyphen in the middle, while the general versions use a dot.

[New Model] The strongest open-source model glm-5.2 is now available, with performance close to claude-4.8-opus and excellent cost-effectiveness. The king of domestic coding—welcome to try it!

[Payment Notice] Sorry everyone. Alipay and WeChat payment channels across the web are under risk control and temporarily under maintenance. The card network is open as a backup. You can go to https://pay.ldxp.cn/shop/lotte-library to place orders via Alipay and get direct recharge codes. As an apology, the card network is running a 9–9.5 discount promotion for one week. If you need WeChat Pay, please contact customer service QQ: 746999675, and the discounted price also applies. We'll notify you as soon as the direct website recharge is restored.

[Payment Notice] The Alipay payment channel is under maintenance today, expected to complete by 14:00 on June 7. The WeChat payment channel has been under maintenance due to recent official network fluctuations. For WeChat or Alipay payments, you can also contact customer service QQ: 746999675 directly. The customer service always gives extra bonuses to returning customers. Thanks for your understanding and support!