DDchat AI vs VibeBang

DDchat AI is a polished, minimalist native iOS AI client with web search, multimodal understanding, and HTML rendering. Connect your own APIs to use different providers and models in one flexible, fluid app. — BEAUTIFUL BY DESIGN — From rounded bubbles to smooth streaming animations, from live scrolling reasoning previews to syntax-highlighted code blocks, every pixel is polished. Full Light and Dark Mode support, liquid glass, and gradient accents create a truly native experience. — OFFICIAL APIs AND THIRD-PARTY GATEWAYS — One app to connect with advanced AI models from around the world. Use DeepSeek, Kimi, MiniMax, Zhipu, MiMo, Qwen, and more. Connect to official providers or compatible third-party gateways, choose your provider, and switch models to suit each task. — AUTOMATIC MODEL CAPABILITY DETECTION — A built-in capability database detects whether a model supports advanced reasoning, vision, tool calling, and more, then applies the appropriate parameters. The database updates silently from the cloud, helping new models work without an app update. — HTML RENDERING ENGINE — HTML tags in AI replies can be rendered directly in the conversation, making cards, tables, lists, and other rich content easier to read. Built-in Markdown parsing and syntax highlighting for 10+ languages, including Python, Swift, Go, and Rust, keep code and quotes clear as well. — MATH AND DIAGRAM RENDERING — Render LaTeX equations and Mermaid diagrams right in your conversations, from formulas to flowcharts and sequence diagrams. Zoom in on diagrams and save them as images for study, organization, or sharing. — WEB SEARCH WITH SOURCES — Use built-in model search or connect a range of independent search services. Models can search and browse the web based on your question, with citations and sources shown directly in the answer so important information is easy to verify. — MULTIMODAL IMAGE UNDERSTANDING — Take a photo or select multiple images from your library and chat directly with vision-capable models. Analyze screenshots, understand scenes, or read text from images with ease. — VISUALIZED REASONING — See a model's reasoning as it is generated, then expand it for more detail. When supported, choose different reasoning levels to balance response speed and depth. — TASK-SPECIFIC ASSISTANTS — Create dedicated AI assistants for different scenarios, such as translation, coding, writing, or research. Each assistant has its own system prompt, sampling parameters, context length, and default model. — ASSISTANT CREATION AND OPTIMIZATION TOOLS — Describe what you need in one sentence, and AI can create a new assistant for you. If the results are not quite right, use another request to refine its prompts and related settings so it works the way you want. — PRIVACY-FIRST, STORED LOCALLY — API keys, chats, and assistant settings are stored locally on your device. Requests go directly to the AI provider you choose, without passing through a DDchat intermediary server. No analytics, no ads, no tracking. — MORE HIGHLIGHTS — - Multi-conversation history grouped by assistant - Fetch model lists online or add models manually - Duplicate, import, and share assistants - Keep streaming replies running in the background and generate with multiple assistants at once Getting started: 1. Add an API provider in Settings, either an official endpoint or a compatible gateway 2. Create an AI assistant and configure its prompt and parameters 3. Choose a model and start chatting Note: This app requires you to provide your own API keys. It does not include or provide any AI service. — CONTACT — For feedback or questions, please email: ddchatsupport@gmail.com

VibeBang 是一款 100% 本地离线的语音输入工具,基于开源 X-ASR 中英双语流式引擎驱动。说话即出字,随手就能改,说完直接送到你的工具里。 【为什么选 VibeBang】 • 本地离线:语音与文字全程在本机处理,绝不上传,断网也能用。 • 中英文自由说:中英混说无缝切换,实时上屏。 • 流式识别:边说边出字,反馈快。 【两种说法】 • 按住说:按住麦克风即说、松开即停,一段一段来。 • 点一下持续说:点一下开始、再点停止,免手连续口述。 【自定义热词,识别更准】 • 把人名、术语、品牌、专有名词加进词典,识别更倾向这些词;分页编辑、强度可调。 【智能修正】 • 同音字纠正:把「贾阳清」自动改回你词典里的「贾扬清」。 • 文本替换 & 口令:错形一键换正;说一个触发词,自动展开整段邮箱 / 签名 / 模板。 • 数字规整、去口水词等可选项,默认不动你的原文,想要再开。 【随手编辑】 • 轻点一句直接改字;长按「重新说」用语音重录替换,弹框确认不误覆盖。 • 大爆炸:原地把一句话「炸」成词块,点选 / 滑动多选,复制、删除、键盘改写、语音改写,逐词精修。 【持续候机(OnCall)】 • 全屏免手听写,每说完一句自动落成一张卡片,可逐条复制。 【一键发送】 • 把识别结果直接送到 Termius、Blink、a-Shell、ShellFish、MS 远程桌面、ToDesk、向日葵、RustDesk、VS Code、Working Copy 等工具,或复制到剪贴板。点了只复制 + 打开 App,你手动粘贴,绝不替你操作。 【历史 · 日历】 • 每次听写按会话自动存档;月历热力图一眼看出哪天说得多,点某天回看当天记录,可搜索、可逐句修改。 【本地共享(进阶)】 • 可在本机开一个只读接口,把你的语音记录 / 词典安全接给自己的 AI 编程助手;默认关闭、仅本机、需密钥。 【贴心细节】 • 深色 / 浅色一键切换;字号可调;撤销 / 重做;长按防误触清空;支持蓝牙耳机麦克风。 隐私优先,本地优先。VibeBang 让语音输入既快又稳,还放心。