Änderungsprotokoll & Produktaktualisierungen

DeepSeek V4.1 Flash is the updated version of DeepSeek's fast model, replacing V4 Flash (0731). It is available on the free plan.

GPT OSS 120B is OpenAI's open weight model, a 120B parameter mixture of experts with adjustable reasoning effort. It is free to use in Aymo AI.

Grok Build 0.1 is xAI's fast coding model, trained specifically for agentic software engineering. It is built for interactive coding agents, tool use, and multi-step development work rather than general chat. It accepts text and image inputs, produces text output, and has a 256,000-token context window.

GPT-5.6 Sol is OpenAI's top-tier model in the GPT-5.6 family, above Terra and Luna. It is built for the hardest reasoning, coding, and analysis work.

Gemini 3.8 Flash is Google's newest Flash model, replacing Gemini 3.7 Flash. It handles coding, reasoning, and agent tasks at Flash speed.

Claude Fable 5.1 is Anthropic's most capable model tier, above Opus. It adds additional safety measures for biology, cybersecurity, and AI research.

Kimi K2.7 Code is Moonshot AI's coding-focused model. It is tuned for writing, debugging, and refactoring code across long development tasks.

Gemini 3.7 Flash is Google's newest Flash model, built for coding, reasoning, and agent tasks while keeping Flash-level speed.

DeepSeek V4 Pro (0813) is DeepSeek's newest flagship model, built for reasoning, coding, long documents, and agent tasks. It supports a 1M-token context window.

Grok 4.6 is xAI's newest flagship model. It performs well on questions that draw on real-world knowledge, and handles long-running coding tasks better than the previous version.

Aymo AI now generates images from text. Write a prompt, choose a model, and generate. You can then edit the result in the same place: remove the background, upscale it, expand the frame, or adjust the lighting.

Key Features:

• Five image models: Grok Imagine Image, Seedream 4.5, GPT Image 2, Nano Banana 2, and Nano Banana Pro
• Run one prompt across models: send the same prompt to more than one model and compare the results
• Edit in place: remove backgrounds, upscale, expand the frame beyond the original image, and adjust lighting without generating again
• Style presets: realistic, 3D, anime, illustration, pixel, minimal, cinematic, and product photo, or write your own style description
• Format control: set aspect ratio and resolution before generating
• One subscription: covers image generation, chat, and editing

Seedream 4.5 is ByteDance's image model, now in Aymo's image generator. It supports text-to-image and image-to-image, and you can start using it on the free plan.

Nano Banana Pro is Google's higher-capability image model, now in Aymo's image generator. It supports text-to-image and image-to-image generation and outputs at up to 4 K resolution.

Nano Banana 2 is Google's image model for everyday generation, now in Aymo. It supports text-to-image and image-to-image, and you can start using it on the free plan.

Grok Imagine Image is xAI's image model, now in Aymo's image generator. It supports text-to-image and image-to-image, and you can start using it on the free plan.

GPT Image 2 is OpenAI's image model, now available in Aymo's image generator. It handles text-to-image and image-to-image, so you can generate from a written prompt or supply a reference image and change it.

We just raised the monthly credit limits on both paid plans.

Premium goes from 2K to 3K credits per month. Business goes from 6K to 7K.

Same price on both. Nothing to do on your end, the new limits are already live on your account and apply to your current billing cycle.

Qwen3.8-Max is Alibaba's flagship Qwen model. It is a 2.4T-parameter mixture-of-experts model with native multimodal input and support for very long context.

DeepSeek Flash has been updated to DeepSeek-V4-Flash-0731 - This is our new default when you use DeepSeek Flash V4

DeepSeek V4 Flash (0731) is the updated version of DeepSeek's fast model. It is available on the free plan.

AI Copywriter is a new tool for writing headlines, ads, product descriptions, and emails. It draws on multiple models so you can generate several versions of the same copy, compare them, and refine the one that fits your brand voice.

GPT-5.6 Terra is OpenAI's balanced model, offering strong capability at a lower cost than the flagship tier. It supports reasoning, web search, and image input, and works well for coding, analysis, and writing.

GPT-5.6 Luna is OpenAI's fastest and most cost-efficient GPT-5.6 model, built for high-volume workloads. It supports reasoning, web search, and image input, and handles coding, analysis, and writing.

Gemini 3.6 Flash is Google's new default Flash model. It reads text, images, video, audio, and PDFs, and handles coding and multi-step tasks.

Gemini 3.5 Flash-Lite is Google's fastest and least expensive Gemini model, built for high-volume tasks such as search, translation, and document processing.

Claude Opus 5 is Anthropic's newest Opus model for complex work, with an adjustable effort control and output that checks against itself before replying.

Kimi K3 is Moonshot AI's open-weight flagship, with native vision, a 1-million-token context window, and reasoning on by default. It is strong at coding, long-context work, and knowledge tasks.

Aymo AI Chrome Extension is now available on the Chrome Web Store. This is the first version, bringing Aymo into your browser so you can use AI on any page without switching tabs.

Key Features:

• Works on any page: Use Aymo from wherever you're already browsing
• In-browser access: use AI help without opening a separate tab or window
• Free install: available now from the Chrome Web Store

Grok 4.5 is xAI's smartest model to date, built for coding, agentic tasks, and knowledge work. It has built-in reasoning and live web search, which makes it useful for deep analysis and research-heavy chats.

Tencent Hy3 is a cost-efficient mixture-of-experts model for coding, reasoning, and agent tasks. It supports a 256K token context window and answers quickly on simple prompts while reasoning more on harder ones.

Claude Sonnet 5 is Anthropic's newest Sonnet model, built for agentic coding, reasoning, and analysis. It delivers accuracy close to Opus-class models while staying faster and lower cost, and supports reasoning, vision, and file context.

Prompt Library gives you a set of ready-to-use prompts for common tasks, plus a place to save and organize your own. Run any prompt with any model in a single click, so you stop rewriting the same instructions.

Key Features:

• Curated prompts: browse a library of prompts covering a wide range of tasks
• Save your own: store prompts you use often and reuse them across threads and projects
• Organize your way: group and arrange prompts to match how you work
• One-click run: send any prompt to any model without copying and pasting
• Search and filter: find prompts by keyword or by compatible model

Compare mode runs the same prompt across multiple AI models simultaneously and displays the results side by side. Read them together, then keep the one that fits your task.

Key Features:

• Run once, get several answers: send a single prompt to multiple models in one go
• Side-by-side view: read the responses next to each other instead of switching between chats
• Keep what works: carry the answer you want forward and drop the rest
• Judge before you commit: see how models differ on reasoning, writing quality, and accuracy for your specific task

GLM-5.2 is Z.ai's flagship model for long-horizon tasks, with a usable one million token context window. It excels at long, autonomous work sessions and complex development challenges.

MiniMax-M3 is a multimodal model that understands images and text, built for long and complex conversations. It suits demanding coding and analytical work.