Moltis Review 2026: The Rust Agent Server That Runs Forever on Your Hardware

TL;DR

📊 At a Glance

Moltis is the most practical open-source agent server for multi-channel deployment. It wins on setup speed and channel support, but trades ecosystem breadth and production stability for that simplicity.

Most AI agents are either cloud-dependent or require extensive setup. Moltis takes a different approach: a single Rust binary that runs on your hardware, remembers everything, and reaches you through every messaging channel you use.

It's not a chatbot wrapper. It's a persistent agent server — always on, always local, secure by design.

Here's the full review.

What Is Moltis?

Moltis is an open-source (MIT), persistent personal agent server written in Rust. It's a single binary that:

• Runs on your hardware (Mac, Linux, Windows, Raspberry Pi)
• Connects to 20+ LLM providers
• Reaches you through 8+ messaging channels
• Remembers everything with persistent memory
• Executes code safely in sandboxed environments
• Extends itself with runtime-created skills and MCP tools

Created by Fabien Penso, Moltis is labeled as alpha software but already has an active community on GitHub and Discord.

🎯 Who Should Use Moltis

Moltis is for anyone who wants a personal AI agent server that works across all their messaging platforms without cloud dependency or complex setup.

• Multi-channel communicators — who want one agent across Telegram, Discord, Slack, WhatsApp, Matrix, and more with shared context
• Privacy-first users — local-only operation with no cloud relay, keys and data never leave your machine
• OpenClaw migrants — built-in import reads your existing OpenClaw config, skills, and identity files
• Raspberry Pi / homelab enthusiasts — single static binary runs on low-power hardware

If you need 5,700+ skills and production stability, OpenClaw is safer. If you want the fastest path to a multi-channel persistent agent, Moltis delivers.

Pros & Cons

🔬 Detailed Analysis

📋 Score Breakdown

Overall ToolBrain Score: 7.5 / 10

🚀 Setup: 60 Seconds

I timed it:

class="language-bash">curl -fsSL moltis.org/install.sh | sh

That took 12 seconds. Then:

1. Open http://localhost:13131 in a browser
2. Paste an API key (I used an OpenRouter key for multi-model access)
3. Pick a model

Total time from start to chatting: 47 seconds.

No Node.js. No npm install. No Docker. No dependency hell. One binary, one config screen, done.

Channel Support: Everywhere at Once

Moltis's channel support is its standout feature. It natively connects to:

Channel	Status	Use Case
Web UI	Built-in	Desktop management, configuration
Telegram	Native	Day-to-day messaging
WhatsApp	Native	Mobile interactions
Discord	Native	Community management
Slack	Native	Work communication
Matrix	Native	Decentralized chat
Nostr	Native	Protocol-native messaging
Microsoft Teams	Native	Enterprise communication
iOS	Soon	Mobile native app
GraphQL API	Built-in	Custom integrations
JSON-RPC	Built-in	Developer tooling

Each channel connects directly — no webhooks, no middleware, no cloud relay. Your agent is present everywhere without your data passing through a third party.

I connected Telegram and Discord in about 3 minutes total. The agent responded in all channels with shared context and memory.

Security Model

Moltis takes security seriously:

• Sandboxed by default — the agent can't touch your filesystem unless you explicitly allow it
• Local-first — your keys and private data never leave your machine
• Code execution — runs in Docker or Apple Containers with resource limits
• SSRF-protected — web browsing prevents server-side request forgery attacks
• No cloud relay — channels connect directly without a middleman

This is a stronger default security posture than OpenClaw, which requires manual configuration for sandboxing.

LLM Provider Support

Moltis supports over 20 providers out of the box:

Cloud: Anthropic, OpenAI, Google Gemini, Mistral, DeepSeek, Fireworks, Cerebras, MiniMax, Moonshot, Z.AI, Venice

Local: Ollama, LM Studio, Local GGUF (CPU/GPU), MLX (Apple Silicon), Hugging Face

Meta: OpenRouter (access to 200+ models through one key)

OAuth (zero config): GitHub Copilot, OpenAI Codex, Kimi Code

You can switch providers per-session or use different models for different channels. I used OpenRouter for primary inference and Ollama for local fallback — switching required a config change and a restart, no code modifications.

Persistent Memory

Moltis maintains durable session state with vector + full-text search. It remembers across sessions, channels, and restarts.

I tested this by asking a question about a project on Telegram, switching to Discord and asking a follow-up, then restarting the binary and asking again. Moltis remembered the context in all three cases.

The memory is stored locally in SQLite with embedding vectors generated by your chosen LLM provider. No external vector database required.

OpenClaw Import

Moltis recently added a full OpenClaw import feature. You can migrate your existing OpenClaw configuration, skills, and identity files with a single command:

class="language-bash">moltis import openclaw

This reads your OpenClaw config, converts the provider and channel settings, and imports your SOUL.md / IDENTITY.md files. I tested it with a basic OpenClaw setup and it worked — though complex skill configurations needed manual adjustments.

💰 Pricing

Where Moltis Falls Short

🔄 Alternatives

Feature	Moltis	ZeroClaw	OpenClaw	Vellum
Language	Rust	Rust	TypeScript/Node	Unknown
Binary size	~15MB	3.4MB	~390MB	N/A
Setup time	47 seconds	5 minutes	30+ minutes	2 minutes
Channels built-in	8+	Via skills	Via skills	Web/CLI/iOS
Sandboxed default	✅	✅	❌ (manual)	✅
OpenClaw import	✅	✅	N/A	❌
Memory	Vector + full-text	SQLite/vector	Plugin-based	8-type memory
Price	Free (MIT)	Free (MIT)	Free (MIT)	Free / $15/mo
Maturity	Alpha	Pre-1.0	Stable	Early

Moltis wins on channel support and default security. ZeroClaw wins on binary size. OpenClaw wins on ecosystem breadth.

❓ FAQ

← Back to all posts