Hope Agent·v0.60.0·Local-first · Open source · MIT

An AI assistant that gets work done and remembers you.

Remembers you, grows over time, pursues goals autonomously, orchestrates work dynamically, and stays reachable from every chat app you use. The native installer includes templates for major model providers—paste an API key or sign in and start chatting, with no runtime, CLI, or config wrangling.

Hope Agent preparing the Kanso launch pack with project conversations, a deliverables table, and floating session status
From brief to deliverables · Live UI, demo projectWatch the 3-minute tour
50Provider templates
481preset model entries
12IM channels
12UI languages
Product tour

Watch Hope Agent in 3 minutes

English narration and captions. Explore goals, tools, memory, knowledge, and design in one desktop workflow.

Product presentation built with real Hope Agent components and fixed demo data.

Open video

See how Hope Agent gets work done

Explore the real interface, from a Kanso launch brief to design, connected knowledge, and usage insights.

Hope Agent preparing the Kanso launch pack with project conversations, a deliverables table, and floating session status

From a project brief to three deliverable drafts, with project history and session status in view.

View full size

1920×1080 screenshots of the running Web GUI. Projects, sidebar entries, and notes use demo content; deliverables come from actual runs. Dashboard metrics use illustrative sample data.

Features

Completes work, not just replies

Side chats, tabbed panels, scheduling, and autonomous execution connect design, knowledge, and cross-device work to deliver results.

Goal · keep pursuing the outcome

Give Hope Agent an outcome and completion criteria, and it keeps decomposing, executing, checking, and advancing—with budgets, pause/resume, and a conservative evidence-backed completion audit.

Workflow · dynamic orchestration

The model organizes phases, conditions, parallel work, multiple agents, tools, diffs, reviews, and verification as needed, with durable pause, resume, cancel, and conservative recovery.

Loop · continue on time or events

Continue on intervals, conditions, internal events, or model-selected wakeups. Each iteration can resume the chat or launch a Goal-bound Workflow, with budgets, backoff, and no-progress protection.

Side chats & tabbed panels

Ask a separate question with /side while the main task runs. Reference past chats with @ or import read-only local Codex history. Files, diffs, plans, and the browser share reorderable right-side tabs.

One core, three ways to run

Desktop app, headless service with an embedded Web GUI, or an IDE backend — the same pure-Rust core underneath.

1

Desktop GUI

How to start

Double-click the app / pnpm dev:desktop

The most complete entry point: full GUI plus an embedded HTTP/WS server, so the desktop can serve remote clients while you use it.

2

Server + Web GUI (HTTP/WS)

How to start

server start subcommand; server install registers a launchd / systemd service

Headless always-on daemon for IM channels and cron jobs; the React frontend is rust-embed-baked into the binary, so opening http://<server>:port in any browser gives the full Web GUI — phone, tablet, or any computer can connect directly, no client install.

3

ACP (stdio)

How to start

acp subcommand

IDE integration — any ACP-capable editor can call Hope Agent as its agent backend.

Batteries included

Provider templates, messaging channels, and interface languages — ready on day one.

50Provider templates·481preset model entries
International
AnthropicOpenAICodexGitHub CopilotGoogle GeminiOpenRouterRequestyAzure OpenAIGroqTogether AIFireworksNovitaPerplexityxAI GrokMistralCohere
China
DeepSeekMoonshot (Kimi)QwenDoubao (Volcengine)Z.AI (GLM)MiniMaxXiaomi MiMo
Local
OllamaAny OpenAI-compatible endpoint

12IM channels

TelegramDiscordSlackFeishuGoogle ChatLINEQQ BotSignaliMessageIRCWeChatWhatsApp

12UI languages

Simplified ChineseTraditional ChineseEnglishJapaneseKoreanSpanishPortugueseRussianArabicTurkishVietnameseMalay

Open it, paste an API key, start chatting

One native installer. No runtime to set up, no command line to learn.