View the latest product updates
If you are unsatisfied with Advanced Task or client-generated task results, you can apply for credit compensation. Please upgrade the client to 0.1.15 or later. We usually reply within 3 business days. Read the details
GPT 5.6 Sol, OpenAI's strongest model in the 5.6 series, is now live in Chat, Advanced Task, and the Desktop App. Chat reasoning defaults to High and can be customized anytime.
Chat, Advanced Task, and Image Generation are now together on the web. Limited-time credit discounts for Advanced Task: 20% off on web, 80% off on desktop.
The desktop app now ships with the latest models GPT 5.6 Sol and Claude Fable 5, plus built-in skills for document processing, data analysis, poster design, coding, and more. Use it in the client and enjoy 90% off credits for one week.
New “converse with famous figures” skills are here. Chat one-on-one with history’s greatest minds—Charlie Munger, Qian Xuesen, Richard Feynman, and Karl Popper—and learn from their wisdom.
Claude Fable 5 is live—a Mythos-tier model built for automation work and coding. Fable 5 and Opus 4.8 are available together in chat and Advanced Task. Enjoy 25% off for one month, for a limited time.
New recreate mode lets you replicate with one tap—no prompt needed. Browse tens of thousands of free examples for more inspiration, from academic illustrations and portrait photography to creative posters. Visit Design Explore to get started.
Owl Composer is built for feminine-focused AI companionship and fiction writing, with a minimal, private, immersive text space—ideal for free-flow, stream-of-consciousness creation.
GPT Image 2 now supports generating multiple images in one request—just say how many you want in your prompt. Multi-image credits are charged per image at the same rate, so more volume without paying extra; batches may take a bit longer.
Draw directly on images to mark exactly what to change. The model recognizes annotated areas so it understands your intent. Combine marks with text prompts to edit faster.
GPT 5.5 and DeepSeek V4 Pro are now available. Both perform well on reasoning, writing, and coding tasks. Once API stability testing is complete, they will be added to Advanced Task — ready to use with cloud-based agents.
Temporary chats are not saved and are not added to memory—ideal for one-off or quick questions with less overhead. Open when you need them; more private and flexible.
Advanced Task delivers complete Word, PPT, code, and other files directly; supports long-running background jobs with strict adherence to complex instructions and no missed key details. 200K context; limited-time 80% off credits at moely.ai/advanced-task.
Claude Opus 4.7 is live at the same price as Opus 4.6. It is similar to Opus 4.6 for simple tasks like everyday chat, and performs better on complex multi-step tasks—we recommend pairing it with advanced workflows on app.moely.ai.
Claude Opus 4.6 and Grok 4.2 are now available. Claude Opus 4.6 is currently 50% off for one week promotion, and pricing now explains plans more clearly with standalone credit purchases.
File processing is fully optimized for images, PDF, Excel, Word, PPT, audio, and video. Free users: context window expanded from 8K to 16K; paid users: from 32K to 64K.
GPT-5.4 is live, unifying GPT and Codex into one system with stronger coding, document understanding, and tool use.
Added GPT-5.3—the Instant mode under ChatGPT subscriptions. New Nano Banana 2 image model for higher-quality output.
Gemini 3.1 Pro is live with stronger reasoning, multimodal understanding, and stability. Free trial credits included.
New vertical search for academic, news, social, and developer sources—integrated with Google Scholar, GitHub, Reddit, and more for cross-source verification and more trustworthy results.
Create and edit images with a top-tier AI image model. Supports 1K, 2K, and 4K.
This page currently shows major updates from 2026 only.