Sencha: LLM Chat Client

Sencha is a clean, native client for the AI access you already have. Connect OpenRouter, OpenCode, direct provider APIs, or compatible custom endpoints – then chat with your preferred models in one fast interface. No app subscription or in-app purchases. Bring Your Own AI Access • Connect OpenRouter, OpenCode Go, OpenCode Zen, OpenAI, Anthropic, Google AI Studio, NVIDIA, and Amazon Bedrock. • Add compatible OpenAI or Anthropic endpoints, including servers you run yourself. • Keep multiple connections for the same provider, such as Personal and Work. • Search models, pin favorites, filter free models, and review pricing, context limits, and supported features. • View token usage and estimated cost for individual responses when the provider reports them. • Use prompt caching with supported providers and models. Work With Documents and Images Attach photos, PDFs, text files, source code, Markdown, CSV, JSON, DOCX, and more. For each PDF, choose the approach that fits the document: • Send the original PDF to a compatible model. • Extract its text on your device. • Run on-device OCR for scans. • Send pages as images when charts, diagrams, or layout matter. Before you send a PDF, review its page count, file size, available text, and estimated token count. DOCX files are converted to Markdown on your device. Built-In Web Tools Give compatible models built-in Web Search and Web Fetch without setting up an additional search API key. Web Fetch can extract readable page content directly on your device, with a configurable fallback for sites that block direct access. You can also use native web search, web fetch, and code execution from providers and models that support them. Made for Working With Answers • Select text naturally across paragraphs, lists, tables, and code. • Read long responses, syntax-highlighted code, formulas, and wide tables without cramped formatting. • Copy responses as rich text or Markdown, or reformat them for Slack, Telegram, and WhatsApp. • Copy a complete table as rich text, Markdown, or an image. • Choose reasoning modes and effort levels supported by each model, and show or hide reasoning traces. • Search your chat history and stop or retry responses at any time. • Customize a system prompt with dynamic date, time, and time zone variables. Flexible Voice Input Dictate with Apple Speech or downloadable speech recognition models that run on your device. Cloud transcription through OpenAI or a compatible Whisper endpoint is also supported. Private by Design Sencha has no developer-operated backend. Requests go directly from your device to the service or endpoint you choose. • API keys are stored in the device-only data-protection keychain. • Chat history stays in a local on-device database. • Photo metadata, including location data, is removed before upload. • PDF text extraction and OCR run on your device. • Apple Speech and downloaded speech models can keep audio on your device. • The on-device web reader can fetch and extract pages without a third-party reading service. Built for Apple Devices Use the same native experience on iPhone, iPad, and Mac, with Dark Mode, Dynamic Type, iPad split view, external keyboard support, and multiple windows on Mac. Free, With No App Subscription Sencha is free to download and use, with no app subscription or in-app purchases. AI provider charges or subscriptions may still apply, and you need access to at least one supported service or compatible endpoint. Sencha is an independent client and is not affiliated with or endorsed by the third-party providers named above.