Announcement2026-09-14 07:56
Concurrency Limits Lowered for GPT-PRO and CC-MAX Groups
September 14 Update
Because OpenAI and CC-MAX accounts currently do not allow multiple concurrent connections, and it has been discovered that users are maliciously making high-concurrency calls to the gpt-5.6-luna model, the concurrency for individual GPT-PRO and CC-MAX groups is now limited to 30 RPM to meet personal usage needs. Studios or users with higher call volumes may contact us to negotiate an increase in concurrency.
---
Resolution
The concurrency for individual GPT-PRO and CC-MAX groups has been adjusted to 30 RPM.
Announcement2026-09-08 22:57
GPT Official Self-Service Subscription Feature Is Now Live
September 11 Update
**Current Status**
Officially, the Pro 20x top-up entry has been temporarily closed, and GPTEAM has synchronized with the official move to temporarily close the 20x top-up entry.
---
September 9 Update
GPT Official Self-Service Subscription Service is now live. You can go to "Purchase Plans" and select "Official Subscription" to place an order yourself, supporting top-ups for Plus, Pro 5x, and Pro 20x. The entire process is automatic with no manual intervention, and you can place an order at any time. The top-up is made to your own ChatGPT account, with the subscription completing in about 3-5 minutes and the top-up process finishing in 5-8 minutes, provided by GPTEAM.
**Service Description**
- The subscription uses direct top-up via Philippine region cards, not IOS subscription or Google Pay payment.
- Legitimate card top-up · 30-day subscription warranty.
- Fully automated top-up; after self-service ordering, the system automatically activates it.
**Warranty and Usage Reminders**
The warranty only covers subscription benefits; account bans are not covered by the warranty. Please comply with OpenAI's official usage rules, do not engage in cracking restrictions or abuse, and there is basically no situation of being banned by the official platform.
Announcement2026-09-08 22:57
GPT Official Self-Service Subscription Is Now Live
September 9 update
GPT official self-service subscription is now available. You can go to "Purchase Plans" and select "Official Subscription" to place your order. It supports top-ups for Plus, Pro 5x, and Pro 20x. Fully automatic top-up with no manual intervention; you can order at any time. The top-up goes to your own ChatGPT account. Subscription completes in about 3-5 minutes, and the top-up process finishes in 5-8 minutes. Service provided by GPTEAM.
**Service Description**
- Subscription is a direct top-up via Philippine-region card, not an iOS subscription or Google Pay payment.
- Regular card top-up · 30-day subscription warranty.
- Fully automated top-up; the system activates automatically after you place a self-service order.
**Warranty and Usage Reminder**
The warranty only covers subscription benefits; account bans are not covered. Please follow OpenAI's official usage rules and do not attempt to crack restrictions or abuse the service. Basically, there is no risk of official account banning.
Announcement2026-09-07 11:08
【GPT‑6 使用提醒】请暂缓使用，GPT‑5.6-sol 不受影响
9 月 7 日更新
我们当前提供的 GPT‑6（gpt-6-astra）存在严重的回答质量下降问题。为避免影响任务效果、浪费调用额度，建议大家暂时不要使用，尤其是编程、智能体和复杂推理任务。
GPT‑5.6-sol 不受此次问题影响，可以继续使用。请检查客户端和自动化任务中的模型配置，按需切换。
**当前状态**
技术团队正在处理 GPT‑6 的质量问题。恢复并通过验证后，我们将第一时间发送通知。
**处理方案**
如近期确需使用 GPT‑6，建议通过官方账号订阅 ChatGPT 会员，在官方渠道使用。如需订阅协助，可联系我们：提供 20–200 美元档位的订阅服务，具体方案请咨询。
对于由此带来的不便，我们深表歉意，感谢理解与支持。
Announcement2026-09-07 11:08
[GPT‑6 Usage Notice] Please temporarily hold off on using it; GPT‑5.6-sol is unaffected
September 7 update
The GPT‑6 (gpt-6-astra) currently provided by us has a serious issue of degraded response quality. To avoid affecting task outcomes and wasting usage quota, we recommend that you temporarily refrain from using it, especially for programming, agents, and complex reasoning tasks.
Models such as GPT‑5.6-sol are relatively less affected by this issue and can continue to be used. Please check the model configurations in clients and automated tasks, and switch as needed.
**Current Status**
The technical team is addressing the GPT‑6 quality issue. Once it is restored and passes verification, we will send a notification as soon as possible.
**Handling Plan**
If you truly need to use GPT‑6 in the near term, we recommend subscribing to ChatGPT membership through an official account and using it via official channels. If you need subscription assistance, please go to purchase plans and choose the official subscription to make a purchase.
We deeply apologize for the inconvenience caused and thank you for your understanding and support.
Announcement2026-09-05 06:57
GPT-6-ASTRA Model Released
Updated September 5
The GPT-6-ASTRA model is now officially available on this platform, with a billing multiplier of 0.04x.
Pricing is as follows:
- Input: $0.4/1M (¥2.692/1M)
- Cache read: $0.04/1M (¥0.269154/1M)
- Output: $2/1M (¥13.458/1M)
Announcement2026-08-06 16:39
GPT Series Model Price Reduction
August 7 update
Codex-Pro series model group pricing has been temporarily reduced from 0.06x to 0.04x. This is a temporary adjustment with no confirmed duration; it may last 1 week, 1 month, or several months, depending on the official risk control situation. Enjoy! Come and use it quickly!
Current status: The price reduction is in effect; the restoration time is yet to be determined.
Announcement2026-08-04 11:21
Claude Series Model Status Update
September 16 Update
To cover the cost of Claude account bans, on September 21, the model group pricing will be increased. The CC-MAX group will no longer offer the fable series models, and the group multiplier will be raised to 0.12x. The fable series models will be offered separately in the CC-MAX-FABLE group (launching September 21), with a group multiplier of 0.15x.
---
September 9 Update
Fable model supply has been restored, and the claude-fable-5-1 model is now available.
**Current Status**
- The price has been adjusted from 0.15x to 0.1x.
- Still a pure-blood Max account pool.
---
August 14 Update
The Claude series model account pool has been restored; Fable-5 will not be launched for now.
**Current Status**
- The price has been adjusted from 0.2x to 0.15x.
- Still a pure-blood Max account pool.
---
August 4 Update
The Claude series model account pool is under maintenance, and the use of groups such as CC-MAX and CC-Kiro is suspended.
**Current Status**
- The CC series model account pool is under maintenance.
- Groups such as CC-MAX and CC-Kiro have been suspended.
The restoration time will be announced separately.
Announcement2026-08-04 11:21
Claude Model Maintenance
August 14 Update
The Claude model pool has been restored; Fable-5 will not be launched for now.
**Current Status**
- Price adjusted from 0.2x to 0.15x.
- Still a pure Max model pool.
---
August 4 Update
The Claude model pool is under maintenance; usage of groups such as CC-MAX and CC-Kiro is suspended.
**Current Status**
- CC model pool is under maintenance.
- Groups such as CC-MAX and CC-Kiro have been suspended.
Restoration time will be announced separately.
Announcement2026-08-04 11:21
Claude Series Model Status Update
September 9 update
Fable model supply has been restored, and the claude-fable-5-1 model is now available.
**Current status**
- Price adjusted from 0.15x to 0.1x.
- Still a pure Max account pool.
---
August 14 update
The Claude series model account pool has been restored; Fable-5 will not be launched for now.
**Current status**
- Price adjusted from 0.2x to 0.15x.
- Still a pure Max account pool.
---
August 4 update
The Claude series model account pool is under maintenance, and access to groups such as CC-MAX and CC-Kiro is suspended.
**Current status**
- CC series model account pool is under maintenance.
- Groups such as CC-MAX and CC-Kiro are suspended.
The restoration time will be announced separately.
Announcement2026-08-04 11:21
Claude Series Model Maintenance
Update on August 4
The Claude model number pool is under maintenance. The use of groups such as CC-MAX and CC-Kiro is suspended.
**Current Status**
- The CC series model number pool is under maintenance.
- Groups such as CC-MAX and CC-Kiro have been suspended.
The recovery time will be announced separately.
Announcement2026-07-29 00:06
SSL Certificate Expiration Resolved
At 02:00 on July 29, 2026, the CloudFlare SSL certificate expired, causing access issues for some services. The certificate has been renewed and services have returned to normal operation.
A new SSL certificate has been issued and deployed, and an automatic renewal mechanism has been configured to prevent similar issues.
Announcement2026-06-25 20:04
June 26 System Update and Risk Control Policy Adjustment Announcement
1. Background and Content Review Launch
OpenAI's official risk-control policies have recently become stricter. We have found that some users attempted to bypass model safety restrictions and generate prohibited content. This behavior directly threatens platform account-pool assets, has caused account bans, and affects normal users.
To keep the platform stable, we have launched the content review system with a built-in intelligent review model.
⚠️ What to do after a block:
If you receive a Cyber Policy warning or trigger a content-review block, stop that topic immediately and continue only in a new conversation window with compliant content.
🚫 2. Strictly Prohibited Content
Any request that attempts to make the model generate, rewrite, complete, disguise, or automate the following content is a violation. Packaging the request as “research,” “security testing,” “role play,” “translation,” or “just ideas” does not make it acceptable.
Limit bypass and safety-policy evasion: Prompts, scripts, or instructions that ask the model to ignore rules, disable safety restrictions, output restricted content, or bypass platform, model, account, or API protections.
Unauthorized network or data access: Unauthorized exploitation, access-control bypass, anti-crawling/rate-limit/paywall evasion, or bulk scraping/export/downloading of restricted API data, private assets, commercial databases, member-only content, or non-public materials.
Credential and account abuse: Obtaining, leaking, testing, stuffing, forging, or abusing passwords, API keys, tokens, cookies, sessions, login states, payment identities, or similar sensitive identity material.
Sexual content and child exploitation: Any sexual or explicit content, with zero tolerance for any content involving child sexual exploitation, abuse, or minors.
Illegal or criminal activity: Assistance, encouragement, or planning for illegal activity, including weapons/explosives, drug trading, hacking instructions, malware/virus code, fraud scripts, money laundering, or regulatory evasion.
Violence and self-harm: Extreme violence, gore, disturbing scenes, or instructions that encourage or enable suicide or self-harm.
Hate and harassment: Discrimination, insults, hate speech, cyberbullying, or malicious harassment targeting protected groups or individuals.
Fraud, privacy abuse, and automation misuse: Content for scams, spam, phishing, information manipulation, bulk harassment, privacy invasion, review bypass, or abusive automated registration/login/order activity.
⚖️ 3. Violation Handling
We will strictly enforce penalties:
Monitoring period: 168 hours, or 7 days.
First penalty: If a user receives 3 confirmed violations within one period, the account will be banned for 7 days.
Permanent ban: If another violation ban occurs after unbanning, the account will be permanently banned with no refund.
🛡️ Review Safeguard
We understand that model review can sometimes be wrong. The system includes a secondary review model. After a block, it reviews the conversation context again:
If it is a false positive, it will not count toward the violation total.
If it is confirmed, violation count +1.
Do not attempt to evade platform rules through word splitting, translation, multi-turn staging, code wrapping, screenshot transcription, or similar methods. A compliant environment protects stable service for everyone.
Announcement2026-06-16 22:57
CC-Max-Test Test Pool Launch Notice
Update on June 20
GPTEAM's new risk control testing pool **CC-Max-Test** is now temporarily online. This group is a pure CC-Max subscription pool, designed as a low-price benefit group for testing risk control. Resources are limited and may go offline at any time. The test is not permanently open; it will be open periodically. Availability probe showing green indicates it is usable.
- Multiplier: **0.1x** (RMB/USD 1:1 multiplier is 0.68x), 2-3 times lower than peer prices
- Example price: Claude Opus 4-8 input price for 1 million tokens is ¥3.386
- Personal PRM = 10
**Current Status**
- Limited resources, **may be interrupted at any time**. Do not rely on long-term stability.
- Prices will remain low during the test period, which will last 7 days to one month. If this plan succeeds, the CC-MAX group price can be further reduced.
If you have any questions, please contact GPTEAM official support. Thank you for your understanding and participation.
Announcement2026-06-16 08:33
Notice on Codex Image-2 Interface Parameter Limitations and Compensation Plan
June 21 Update
After consulting and learning from peers' special algorithms, aligning with peers' 2k and 4k image generation methods, we have basically achieved error-free image aspect ratios. 2k and 4k are revived. Effective immediately, all image generation operations will be temporarily billed at the standard price for 2k resolution.
---
June 16 Update
Due to upstream restrictions, the Codex image-2 interface currently does not support passing resolution and quality parameters, resulting in limited output quality (defaulting to 1k resolution in most cases). To protect your interests, we provide the following solutions:
**Solutions**
- **Exclusive Refund Channel**: If your core business is image generation (image-2 calls account for more than 50% of total calls), you can directly contact customer service to apply for a refund.
- **Exclusive Credit Compensation**: If you choose to continue supporting and using (image-2 call ratio greater than 5%), we will provide corresponding credit compensation based on your historical recharge amount.
- **Billing Standard Adjustment**: Effective immediately, all image generation operations will be uniformly billed at the standard price for 1k resolution.
We are actively connecting with other high-quality image generation channels that support full parameter passing. Once there are new developments, we will release update notifications as soon as possible. Thank you for your understanding and support!
Announcement2026-06-12 11:43
Pricing Notice for Domestic Models
**Pricing Notice for Domestic Models**
Currently, domestic models are still in the testing phase and may experience instability. The final pricing has not yet been determined.
We are actively investigating the following issues:
- Caching issues with some models (e.g., Kimi)
Thank you for your understanding and support.
Announcement2026-06-08 19:38
【Important Notice】GPTEAM Billing System Upgrade and Legacy Asset Value-Added Compensation
Dear GPTEAM developers and users,
Hello everyone.
To further align with international mainstream large model billing standards and provide clearer and more transparent cost calculations for API calls, GPTEAM will undergo a comprehensive upgrade of its platform billing system on **June 9, 2026**.
The core of this upgrade is switching from the original "virtual quota pricing" to "real USD exchange rate pricing." Please rest assured that the sole principle of this adjustment is: **We will never let users who have purchased lose out—not only are historical assets 100% protected, but we also provide excess value-added compensation!**
To help you understand intuitively, **$1** of real quota in the new system is approximately equivalent to **21.3 virtual dollars** in the old system. Although the face value of the quota has decreased, the model deduction rates have been significantly reduced accordingly, so your actual available calls will only increase, not decrease.
Below are the specific asset conversion and compensation plans:
### 1. Lossless Conversion and Value-Added Benefits for Quota Packs
The new system uses real USD pricing and will uniformly adopt the **"old Standard (small quota pack) conversion factor of 0.0470"** that is most favorable to users to convert all historical assets on the platform.
* **Account balance appreciation across the board**: Old balances are converted to new real USD using `original balance × 0.0470`. Because the small pack factor is uniformly applied, users who previously purchased medium or large quota packs (Plus/Pro/Ultra) will see their book assets soar directly.
> **Example**: An Ultra quota pack user originally paid 1199.9 RMB to get 5000 virtual dollars. After conversion, they have 235 real USD. At the new exchange rate, these 235 USD are worth nearly 1598 RMB, directly **appreciating over 30%**.
* **Model rates "skip-level" reduction**: The deduction rates for all base models on the platform have been reduced proportionally or even more. We have deliberately lowered the rates of base models originally close to the Standard tier to a preferential range closer to the original Pro tier. This means that users who previously purchased Standard/Plus quota packs will get more actual calls under the same quota, making their balance last longer.
### 2. Legacy Subscription Plan "Collector's Edition Compensation"
To avoid the system migration affecting users' daily usage, the platform has reset all periodic subscription users' "today's used amount" to zero before the upgrade. Meanwhile, old daily, weekly, and monthly cards will adopt a "collector's edition retention" strategy—not only are benefits unaffected, but daily quotas have actually received a compensatory increase of up to **28%**:
* **Original Daily/Weekly/Monthly Standard**: Adjusted to **$3** per day (approximately 63.9 virtual dollars in the old system), equivalent to **1.28 times** the original daily 50 virtual dollars.
* **Original Monthly Plus**: Adjusted to **$5** per day (approximately 106.4 virtual dollars), equivalent to **1.06 times** the original daily 100 virtual dollars.
* **Original Monthly Pro**: Adjusted to **$7.5** per day (approximately 159.75 virtual dollars), equivalent to **1.06 times** the original daily 150 virtual dollars.
* **Original Monthly Ultra**: Adjusted to **$15** per day (approximately 319.1 virtual dollars), equivalent to **1.06 times** the original daily 300 virtual dollars.
*(Note: The new system will completely remove daily and weekly cards and only offer monthly subscription plans with a new structure. Please check the latest recharge page.)*
### 3. Platform Recharge and New Transparent Pricing Rules
* **Real exchange rate recharging**: After the upgrade, platform quota will be denominated solely in real USD ($). When recharging, settlement will be at a fixed exchange rate (currently approximately **1:6.8**), meaning you pay 6.8 RMB to get 1 USD quota—what you see is what you get.
* **Clear billing formula**: The new rate directly represents "the proportion of platform charges to the official API price."
> **Calculation example**: Assume official gpt-5.5 pricing is $5/M input, which converts to 5 × 6.8 = 34 RMB/M. If the platform sets the model rate at **0.06x**, the actual deduction is: 34 RMB × 0.06 = **2.04 RMB/M**.
> *(If you are used to comparing with the market's "RMB:USD = 1:1" relay station algorithm, simply multiply the platform rate by 6.8. For example, 0.06x at an exchange rate of 6.8 is equivalent to 0.41x at a 1:1 relay station.)*
### 4. Upgrade Time and Notes
The system has already completed asset conversion and updates and is open for recharging. The changes in your book balance are due to the normal exchange rate conversion and value-added compensation mentioned above. Please use it with confidence.
GPTEAM is always committed to providing stable, cost-effective API access services. If you have any questions about balance verification or plans, feel free to contact the admin in the community group for verification.
Thank you for your understanding and support of GPTEAM!
**GPTEAM Operations Team**
**June 9, 2026**
Announcement2026-06-07 17:49
Model Detection Station Inclusion Announcement
The **GPT series** and **Claude series** models on our site have now been included by the following detection platforms:
- [ztest.ai ZhenCe](https://ztest.ai)
- [hvoy.ai Hewei AI](https://hvoy.ai)
Welcome everyone to supervise the quality of our models through the above platforms, ensuring the models are authentic and unadulterated. Thank you for your attention and trust!
Announcement2026-06-04 15:50
Claude Series Models Now Open for Testing!
📢 【New Product Launch】Claude Series Models Open for Testing!
Claude series models (including claude-opus-4-8) are now officially available for testing. This release features the Claude Code Max full-power high-concurrency pool and the pure kiro power account pool (sla99.99% | cctest 100 points), ensuring a stable and high-quality response experience. Everyone is welcome to try them out.
💰 Billing and Multiplier Explanation (using claude-opus-4-8 as an example):
CC-Max Group: Billing multiplier is 5.5x, equivalent to approximately ¥7.422 / 1M Tokens. Temporary multiplier of 4x, reverting to 5.5x on June 8, 00:00.
CC-Kiro Group: Billing multiplier is 3x, equivalent to approximately ¥4.048 / 1M Tokens.
💡 Tip: For specific billing amounts and detailed parameters of each model, please go directly to the 「Model Plaza」. We welcome your experience and valuable feedback!
Announcement2026-06-04 01:47
Codex-Plus Channel Service Adjustment and Future Planning Announcement (Updated June 7)
Dear user, hello!
**June 7 Update**
The Plus gray channel source has disappeared, and the **Codex-Plus** group has been temporarily closed. Please switch to the **Codex-Pro** group to continue using the service. If Plus or Team gray channels become available again in the future, the platform team will strive to secure low-cost access and reopen the Codex-Plus group.
---
**June 4**
Recently, due to OpenAI's extremely strict risk control policies, the gray channels of the Codex-Plus group on this site have largely become invalid. To minimize the impact on your business and usage, we have been continuously deploying high-level accounts from the Pro pool to supplement the Plus group, ensuring that you can still use the related services of this group normally.
**Future Planning**
The technical team will conduct continuous observation and testing of the Plus group channel for approximately one week. If the channel remains unable to bypass risk control and restore stability, in order to ensure the overall service quality of the platform, we will decommission the Codex-Plus group. At that time, this site will concentrate all resources on maintaining and optimizing the core main channel **Codex-Pro**, providing you with a more long-term and stable experience.
Thank you for your continued understanding and support! If you have any questions, please feel free to contact customer service.
Announcement2026-06-03 01:40
OpenAI Official Outage (Updated June 8)
[Announcement] Explanation of frequent errors with OpenAI official models recently
Update on June 8
Crazy Monday is indeed a bit crazy – the official image-2 broke again and cannot generate images 😅
---
Update on June 5
It appears that OpenAI's image-2 generation is having issues; multiple forwarding parties have reported this. Please wait for official fix.
---
June 3, 13:00
Dear friends,
Currently, the official gpt-5.4 and gpt-5.5 models have a high error rate, and this morning gpt-image-2 also showed similar issues.
After our investigation, this problem is an official technical failure from OpenAI, not due to the relay interface. We suggest you take a break or switch to other stable models.
We will continue to follow up on the official fix progress. You can check the service recovery status in real time via OpenAI's official status page:
🔗 https://status.openai.com/
Thank you for your understanding and cooperation!
Announcement2026-05-30 06:32
Notice on Temporary Scheduling of Plus Group Nodes
Dear users,
Hello! Due to recent OpenAI risk control environment impacts, the gray channel resources for the Plus group have become strained. Given that the platform currently has ample Pro resource reserves and today is the last day of the 7-day billing cycle (quota will soon refresh), we have made the following temporary adjustments to fully ensure your calling experience:
Adjustment: Temporarily allocate some Pro accounts to the Plus group to fill resource gaps.
Recovery time: Expected to withdraw the temporarily allocated Pro accounts at 2:00 AM tomorrow, restoring regular routing.
During this period, you will enjoy a better response experience in the Plus group. Pro group users can also temporarily switch to the Plus group to save costs. Thank you for your understanding and support!
Announcement2026-05-19 04:05
New Version Update Announcement
Dear Developers:
We are pleased to inform you that the core functionality part of the new version has been updated! You can now run `npx gpteam` to reconfigure and experience our services.
Since this update involves the mapping migration of the underlying dual databases, if you encounter any bugs, abnormal charges, quota changes, or subscription failures during use, please feel free to contact customer service. We will investigate and resolve them promptly.
In addition, due to the reset of OpenAI quotas two days ago, our promotional event has been postponed. To thank you for your support, our exclusive discount event will officially start today (May 19) at around 15:00!
If you need any help, feel free to contact customer service. Thank you for your understanding and support along the way!
Announcement2026-05-18 02:49
Service Update Notice (Including Free Offer)
📢 【Important Announcement】Major Backend Service Upgrade and Temporary Compensation Plan
Dear Users,
Hello! To provide you with faster response times and a more stable service experience, our backend system is undergoing a major version upgrade. The new version has successfully passed the testing environment and is now entering the final deployment and testing phase in the production environment (estimated to take 1-2 days).
⚠️ Notes During the Upgrade:
Service Fluctuations: During this period, the production environment may experience frequent interface instability, request timeouts, etc.
Feature Suspension: To ensure data security, the system has temporarily disabled the [New User Registration] and [Top-Up] functions.
🎁 Exclusive Transition Plan for Existing Users (Free Access):
To avoid impacting your normal operations, we have opened a dedicated test environment for regular subscribers (including users who have purchased quota packages or periodic subscription plans):
How to Obtain: Please add the administrator on QQ at 3507577690 and include the note 888 (do not forget).
Claim Process: The administrator will invite you to join a WeChat user group, where the BaseURL and Key for the test environment will be distributed uniformly.
Benefits: This test Key has no concurrency limit (please use it responsibly; do not abuse it). The underlying architecture of the new version has been fully optimized, allowing you to experience faster response times in advance!
⏱️ Schedule:
The test benefit is tentatively valid until May 19 at 12:00 noon. If the production environment testing is not yet complete by then, we will automatically extend the validity of your test Key.
Once the production environment is stable, we will notify everyone immediately in the WeChat group and via platform announcements.
Thank you for your continued understanding and support! We apologize for any inconvenience caused.
Announcement2026-05-11 10:40
Non-Stop Service Maintenance Notice
Dear users,
Our platform is continuously undergoing major updates to business-layer services, and will continue to perform rolling updates on servers without downtime.
- Update method: Servers are updated without downtime, and services remain available throughout the process.
- Error feedback: If you encounter any errors or anomalies during API calls, please contact our customer service team promptly. We will quickly investigate and resolve the issue.
Thank you for your understanding and support.
Announcement2026-05-07 13:13
Price Adjustment Notice (Updated June 1)
Dear Developers,
Update on June 1
Price Adjustment: Codex-Pro group price adjusted to 1.5x. Codex-Plus group price adjusted to 0.8x.
Update on May 22
Price Adjustment Plan: The current plan is as follows (not final, subject to minor adjustments based on market conditions):
1. New low-cost channel: Codex-Plus group
To provide more options for budget-sensitive users, we have launched the Codex-Plus group focusing on low prices.
Features: This group consists of pure Plus accounts, no Pro accounts, does not support the gpt-5.3-codex-spark model.
Stability: Since this group uses unstable low-cost/gray channels, if you encounter disconnections during use, please switch back to the Codex-Pro group at any time.
Pricing Notice: It is currently uncertain whether the Plus low-cost channel will completely disappear in June. If the gray channel persists, the group price will increase to 0.8x starting June 1.
2. Stable channel price adjustment: Codex-Pro group
Pricing Notice: To maintain the high cost of official genuine subscriptions, the price multiplier for the Codex-Pro group channel will increase to 2.0x starting June 1.
Update on May 7
Price Adjustment Notice:
Hello everyone! Recently, many gray channels for Codex API relay have largely become invalid, and the remaining ones are mostly single-month finished accounts. It is expected that within the next month, these channels will face mass closures or forced price increases. Many users who have used low-quality relay stations have experienced frequent disconnections, repeated account replenishments, and concurrent limit issues last month.
In contrast, our site has remained rock-solid throughout April and May without any interruptions due to account pool issues, even without restricting concurrency. This is because our main channel uses official Pro subscriptions, fundamentally eliminating the risk of account bans and frequent maintenance (you are welcome to privately message customer service to verify Pro subscription payment receipts).
Therefore, we have been bearing real costs that are dozens or even hundreds of times higher than our peers, while maintaining the same prices as them. To ensure we can continue to provide high-quality and stable services in the future, we plan to implement price and group adjustments around June 1.
To protect the rights of periodic subscription users, we are notifying everyone one month in advance so that you can plan ahead according to your needs.
Thank you for your understanding, trust, and continuous support! We will continue to provide you with the most reliable service as always!
Announcement2026-05-06 15:02
Developer Incentive Daily Ranking Feature Launched
Dear users:
The Developer Incentive Daily Ranking feature is now live. This update applies to all users of this platform.
- New daily ranking: Users can view the current day's ranking on the account overview page.
- Ranking basis: Rankings are based on the user's daily consumption amount (USD). To prevent arbitrage, the effective consumption for the day must reach 2 times the reward amount.
- Reward rules: 1st place $50, 2nd place $30, 3rd place $20, 4th-10th place $10.
- Settlement time: Automatically settled and rewards distributed daily at 00:10 Beijing time.
This feature is now active. We recommend logging into your account overview page to check your current ranking. If you have any questions, please contact customer service.
Announcement2026-04-28 12:28
GPT-Image-2 Image Generation Model MCP
Dear users:
The gpt-image-2 image generation model MCP is now officially available. You can quickly use this capability in the following ways:
- API Key page: A tutorial for using the gpt-image-2 image generation model has been provided; you can check the specific invocation method.
- One-click configuration: Run the `npx gpteam` command to automatically complete the optimal API interface selection and gpt-image-2 MCP configuration.
This feature has been stably deployed. If you encounter any issues, please contact customer service via the contact information below the system announcement.
Announcement2026-04-18 04:21
Do Not Invite Yourself for Profit!!!
Dear user:
The platform risk control has identified that some users have engaged in self-invitation rebate activities, and past actions will not be pursued. To standardize the invitation commission mechanism, the relevant rules are hereby explained as follows.
Effective commission: After a new user registers through your exclusive invitation link and completes a valid purchase, the commission will be automatically settled into your invitation account.
Risk control mechanism: The system will conduct correlation analysis and professional risk control model detection based on multiple dimensions including but not limited to account information, device fingerprint, network environment, payment fingerprint, and client behavior trajectory.
Violation handling: Self-invitation for self-purchase or any form of abnormal arbitrage is strictly prohibited. Once verified, the relevant commission and withdrawal requests will be rejected.
Penalty measures: The commission for the first violation will not be counted, the account for the second violation will be frozen for 7 days, and the account for the third violation will be permanently banned. Do not take risks; fees involved in violations will not be refunded.
Please strictly abide by the platform policy to jointly maintain a fair environment. If you have any questions, please contact customer support.
Announcement2026-04-16 02:38
GPTeam User Discussion Group
Dear users,
To improve service quality and communication efficiency, we are considering establishing a user discussion group, where you can share experiences or provide feedback on service issues.
Group functions: Users can exchange usage experiences and give product suggestions here.
Issue response: If you encounter any service issues, you can raise them directly in the group, and we will respond and handle them promptly.
How to join: If you wish to join, please add our customer service WeChat, and note the email address you registered on this platform along with your purpose.
Platform services remain stable. If you have any other questions, you can reach us through the in-group customer service or the support channels on the management console.
Announcement2026-04-03 12:12
Return Route Upgrade & Cache Hit Rate Improvement
Dear users:
Our platform has completed a comprehensive upgrade of the server return route to further enhance service stability and access experience.
This upgrade focuses on optimizing the return route, effectively reducing the probability of 5xx errors for direct domestic connections, while improving the cache hit rate. According to statistics, an average of 92.7% of inputs hit the cache, resulting in higher overall service availability.
VPN node suggestion: If you need to use a VPN, we recommend prioritizing nodes in the Hong Kong region.
If you encounter any issues during usage, please contact us through the in-platform customer service channels.
