Announcements

Thank you all for your continued support of this site. It has been three years since our last downtime for infrastructure migration. Over the past year, the costs charged by our existing cloud hosting…

The following models are now supported: https://www.ohmygpt.com/models/cloudflare%3Azhipu%2Fglm-5.3-flash https://www.ohmygpt.com/models/cloudflare%3Azhipu%2Fglm-5.3 Cloudflare AI is currently 95% off…

Introduction Added support for the /v1/decisions API Added support for bytedance:deepseek/deepseek-v4-1-flash-261017. Pricing is the same as bytedance:deepseek/deepseek-v4-1-flash-260910: input and ou…

Introduction The DeepSeek official API is currently experiencing an outage this morning. We recommend using DeepSeek V4.1 Flash provided by ByteDance: https://www.ohmygpt.com/models/bytedance%3Adeepse…

New model support added: gpt-6-astra Additionally, since the 5-hour limit for the CodeX subscription plan was recently restored and risk controls have been relaxed, the CodeX discount channel has been…

Introduction We recently signed an EA agreement with Microsoft Cloud sales, granting us a built-in review exemption for Microsoft Cloud along with some discounts. It has been integrated and tested for…

New Model Support v4fve is DeepSeek's latest multimodal model. Compared to v4f0731, in addition to adding visual modality input, its performance has also seen further improvements. It is rumored onlin…

Introduction Significant Increase in DeepSeek Official Pricing Due to the significant increase in DeepSeek's official pricing and the introduction of peak and off-peak pricing, we have adjusted our pr…

Added support for the following SOTA models: gemini-3.7-flash tencent/glm-5.3 grok-4.6

Recently, due to changes in OpenAI's policies, we are unable to continue providing the discounted CodeX API. The recovery time depends on whether OpenAI's risk control policies will be relaxed. If you…

Due to the recent discontinuation of our partnership with the special discount channel for the Fireworks series, the prices have been restored to their original rates.

Introduction Recently, some users reported format compatibility issues with Claude in the Chat.Completions API. Therefore, we will no longer use our built-in conversion logic, and instead directly req…

Added the following Fireworks models Currently, all models are 50% off, with native support for the Chat.Completions API, Responses API, and Messages API. The original pricing is as follows: Added Res…

Added Google's latest SOTA models: gemini-3.6-flash gemini-3.5-flash-lite

We have recently partnered with Tencent Cloud. The glm-5.2 model provider has been switched to Tencent Cloud and is now offering a 20% discount to all users site-wide New SOTA models added: tencent/hy…

New model support added: gpt-5.6 (alias for gpt-5.6-sol) gpt-5.6-sol gpt-5.6-terra gpt-5.6-luna

We have recently added support for Claude Fabel 5 Sonnet 5, glm-5.2, and Grok 4.5. The new models can be found in the model list. Recently, we received feedback from users utilizing the image generati…

We now support the latest DeepSeek v4 series models, with special pricing available for the pro series. The latest OpenAI models are also now supported: gpt-5.5 gpt-image-2

Claude Opus 4.7 Model Support Added Anthropic's latest flagship model claude-opus-4-7: Model Configuration: Context: 1M tokens Output: 128K tokens Input Modalities: text, image, pdf Supports: Tool Cal…

Due to the recent appreciation of CNY against USD, we have adjusted the number of Coins that can be exchanged for 1 CNY to ensure it stays close to the latest exchange rate. Before adjustment: 1 CNY =…

Today we have restored the CodeX special pricing API, with 100% official direct connection forwarding at a 95% discount. You only need to use npx omgvibe to reconfigure CodeX with one command, or ensu…

As Claude's account suspension mechanism becomes increasingly sophisticated, our Claude Code service is currently unavailable. You can switch to regular-priced Claude, CodeX, or OpenCode + other model…

Due to Claude's extremely strict risk control measures recently, maintenance difficulty and costs have increased, and we are unable to continue maintenance during UTC+9 nighttime hours. As compensatio…

Updated at 19:50 Tokyo Time: CC has been restored --- Added support for the following models: gpt-5.4-mini gpt-5.4-nano Due to the current extremely strict risk control policies from the official side…

Due to the inability of Google Cloud's sales team to continue providing lower discounts recently, in order to ensure stable service for our users, the discount for Gemini models on both Vertex AI and…

We just experienced a core component failure that affected some models and APIs: Claude Code dedicated API CodeX dedicated API Gemini (including Vertex AI and Google AI) Fireworks We noticed this issu…

At approximately 02:55:36 JST on February 12, 2026, due to unknown reasons, one of our Cloudflare Warp proxy services stopped working. The models that rely on this service when our servers forward API…

Today we updated our Stripe integration, restored the Stripe JP payment channel, and added automatic top-up functionality for card binding on both Stripe US and Stripe JP. Card Management: https://www…

Recently, our Claude Code service encountered some issues. Around 18:00 Tokyo time on 2026-01-27, we noticed a surge in Claude Code service usage, which caused partial service overload and resulted in…

At 21:44 Tokyo time this evening, OpenAI terminated API access for our primary account, causing an outage of OpenAI-related API services on our site. We have submitted an appeal to OpenAI and register…

Today we updated the pricing library to support the latest gpt-5.2-codex model, which can be used in Responses and CodeX dedicated APIs.

Due to the unexpected termination of a recently signed contract, we have to adjust the Google AI series discount from 0.6 (60% off) to 0.8 (80% off), which means the price will increase from 60% to 80…

We have signed a new contract and secured more favorable medium to long-term Google Cloud discounts, so we've decided to pass these savings on to our users. Previous Discount: 0.9× (10% off official p…

We recently enhanced the compatibility of the Gemini API, supporting more complete Function Call and API compatibility: [x] chat/completions [x] messages - [x] content - [x] role - [x] "system" (=>sys…

We have recently encountered some difficulties maintaining the Claude Code API. We will continue to maintain it, but availability may not be as reliable as usual, and we will pause maintenance during…

We recently updated and maintained the Stripe dependency version, adding the stablecoin payment feature officially provided by Stripe. When using the Stripe US top-up channel, you will now see the sta…

Due to the recent appreciation of CNY against USD, we have made some adjustments to the exchange rate between CNY and Coins Previously: 1 CNY = 34,000 Coins Now: 1 CNY = 35,000 Coins

Starting at 4:00 AM Tokyo time today, our Claude Code related API was temporarily suspended due to Claude's official pause on registrations. We resolved this issue at 11:00 AM, and the Claude Code API…

Recently, we have noticed that a large number of malicious users are still exploiting the invitation mechanism to obtain invitation rewards, which has caused the maintenance pressure and budget expend…

Update Details: Discounts on All Gemini Series Models We recently partnered with Google Cloud's sales team and obtained some official API discounts. We have now made Gemini series model discounts avai…

Recently, due to stricter risk control measures from Claude's official team, account suspensions have been occurring very quickly. Combined with high user demand, our Claude Code service has been expe…

We have restored Claude Code related services as of 9:30 a.m. Tokyo time on 11/27/2025. We apologize to users who experienced inconvenience due to the temporary suspension of this API service over the…

Temporary notice as of November 26, 2025: Due to Claude's official risk control measures, we are currently unable to register new Claude subscription accounts. Therefore, the CC service on this site w…

We recently launched a new announcement system, which will enable us to publish new announcements on the website more quickly and promptly. Added the gpt-5.1-codex-max model, available exclusively thr…

Promotional Activity Launch and Unexpected Circumstances We launched our new user invitation activity today, where both inviters and invitees can receive a $100 Claude Code resource package reward. Af…