Talaria Agent Client vs Hermes Redde

Talaria is a native iPhone and iPad client for Hermes Agent — the open-source agent you run on your own computer. It is not a chatbot with a server behind it. It talks to your machine, runs your models, and holds your credentials on your device. REQUIRES HERMES AGENT You need Hermes Agent running on a computer you control: github.com/NousResearch/hermes-agent Talaria connects to it over your own network. There is no Talaria account and no service in the middle. If you do not run Hermes, this app has nothing to connect to — open it and tap "Look around with sample data" to see how it works first. EVERY CONVERSATION, EVERYWHERE Sessions you started in the CLI, in the desktop app, or through a chat platform all show up in one list, with previews and the branch each was working on. Pick one up on your phone exactly where you left it. WATCH IT WORK Replies stream in as they are written. Every tool run is there to read — the command, the output, the file it touched — instead of a spinner and a summary. Charts and images the agent generates appear inline. YOU STAY IN THE LOOP When the agent is about to do something destructive, it stops and asks. Approvals, clarifying questions, sudo requests and credential prompts arrive on your phone, and nothing runs until you answer. TALK TO IT Hold the microphone and speak. Recognition happens on your phone whenever iOS has the offline language model installed for your language; where it does not, Apple's speech service transcribes the audio. Choose "On the server" for either direction in Settings to send it to your own gateway instead. SCHEDULED TASKS See what your agent is set to do and when, read what happened on past runs, and edit the schedule, the model and the prompt from your phone. SKILLS, TOOLS AND MEMORY Browse installed skills and turn them on or off. See connected MCP servers and the tools they expose. Read the agent's memory. Browse the working directory. WHAT IT COSTS TO RUN Token counts and spend per model, so you can see what a conversation actually cost. CHOOSE THE MODEL PER CONVERSATION Set the model and the thinking level before you send the first message, or change either mid-conversation. Eight thinking levels, from off to ultra. HOW IT CONNECTS Over HTTPS to a gateway with a valid certificate, or with a prompt to inspect and trust a self-signed one — the same way an SSH client asks about a host key. Plain HTTP is allowed only to addresses that cannot leave a private network: Tailscale, your LAN, or localhost over a forwarded port. Credentials live in the iOS keychain. PRIVACY No account. No analytics. No crash reporting. No server operated by us, because there isn't one. Your data goes to the gateway you point it at — and, if you dictate on a device without the offline language model, the audio goes to Apple for transcription. Nowhere else. Talaria is not affiliated with Nous Research.

Redde is a complete iPhone and iPad client for Hermes, the open-source AI agent. Talk to it or type, watch it think and use tools, approve what it wants to run, and manage everything it does: sessions, projects, profiles, scheduled jobs, and the Kanban board. It also talks straight to any OpenAI-compatible model server when you just want a fast answer. You bring the backend. Redde has no account, no analytics, and no cloud of its own; every request goes only to the server you configured. TALK, DON'T TYPE • Tap the mic, ask, hear the answer. Speech recognition runs on your iPhone; the reply is spoken back with the built-in voice or your own Kokoro server. • Hands-free mode keeps the conversation going until you say "that's all". • Interrupt mid-answer: tap the mic to cut in, or type to steer the reply. • In CarPlay, ask through the car's mic and hear the answer through its speakers. • "Hey Siri, ask Redde…" works from the Lock Screen and AirPods. A Control Center button and Lock Screen widget open Redde listening. • On iOS 27, Siri can message your agent in your own words and read the reply back (opt-in). • A voice orb that moves with your voice and Redde's. Pause, replay, or have any reply read aloud. • "Ask Redde a Question" in Shortcuts returns the answer as text for your automations. SPOKEN PREFIXES Start a voice or Siri message with a word of your choice and Redde does what the word says: drops a text prefix in front, switches the conversation to a model you chose, or both. Say "Opus, review this diff" and the conversation moves to Opus before the message goes. The switch sticks for the follow-ups. Set it up once in Settings → Voice: the word, the ways it tends to be misheard, the prefix, the model. Off until you add a rule, spoken messages only, and anything you type stays exactly as typed. SEE THE AGENT WORK • Watch reasoning as it streams, with tool calls as they happen and a Live Activity for progress. • Approve or deny a risky command, answer the agent's questions, a sudo prompt or a missing secret, all from a notification. • Subagents get their own rows; tap one to follow its transcript, steer it, or stop it. A FULL CLIENT, NOT JUST A MIC • Every conversation lives in your gateway's session ledger: start on the phone, continue on the desktop, Telegram or the CLI. Search, pin, rename, fork, archive. • Sessions grouped by project. Scheduled jobs with delivery targets and blueprints. The Kanban board, live. • Rich replies: Markdown, highlighted code, tables, task lists, images, zoomable Mermaid diagrams, and math. • Pictures and files from the agent arrive ready to view: snapshots, charts, PDFs. • Regenerate, edit and resend, or search a long chat. • Attach photos and files; export chats as Markdown. • Share to Redde from any app: text, links, photos or files land in the composer. • Widgets show the last answer or start listening; the Action Button can open Redde listening. • Switch models mid-conversation from the chat header or with /model. • Skills and tool sets, on or off. Edit the agent's memory and context files in place, and create skills or let Redde draft one. YOUR BACKEND, YOUR RULES • Connect to Hermes through the Hermes Dashboard or the Hermes API, on your LAN, over a tailnet, or behind Cloudflare Access. • Or go straight to any OpenAI-compatible endpoint: llama.cpp, llama-swap, vLLM, Ollama on your hardware, or a hosted provider with your key. • Credentials stay in the Keychain. Speech recognition runs on the phone. • Face ID lock, seven themes with your own colors, app icons and voice orbs, iPad split view, keyboard shortcuts. • No account, no analytics, no tracking. The privacy label says "Data Not Collected" because that is what happens. REQUIRES A Hermes agent (github.com/NousResearch/hermes-agent) or an OpenAI-compatible endpoint. Redde includes no model or hosted service. Hermes is an open-source project by Nous Research; Redde is an independent client, not affiliated with Nous Research.