Octomind is unclaimed —

Octomind

Install agents, not frameworks.·octomind.run

Octomind is an open-source runtime for specialist AI agents, distributed as a single Rust binary that installs in seconds. It ships with 108 pre-wired specialist agents across 27 domains (including legal, medical, finance, engineering, security, and devops) that can be installed and run with one command, each pre-wired with the right model, tools, and prompts. It supports 20+ AI providers, adaptive context compression for long sessions, and dynamic MCP tool loading, and can be run self-hosted or via Octomind Cloud, which provides hosted machines and models.

What it's for

Investigating a question from background, evidence, and counter-argument angles to produce a cited reportTaking a product idea from market exploration to a go-to-market kitParallel multi-lens code review with adversarial verification of findingsTurning one article into platform-native social media draftsTroubleshooting Kubernetes issues (e.g. CrashLoopBackOff)Interpreting blood test resultsChaining multiple specialist agents into a pipeline workflowRunning AI assistance directly inside a code editor via ACPRunning persistent background sessions that react to external events via webhooks

Features 18

108 pre-wired specialist agents across 27 domains
18 pre-built multi-step workflows (e.g. research, launch, deep-review, promote)
ACP (Agent Client Protocol) support for running as a sub-agent in editors
Adaptive, cache-aware context compression (saves up to 72.5% of tokens)
Cross-session learning that reuses lessons across sessions
Daemon mode with webhook hooks for background/event-driven sessions
Developer API access
Dynamic MCP: register new MCP servers at runtime without restart
Hard per-request and per-session spending caps
Mid-session model switching via /model command
Octomind Cloud hosted machines, sessions, and models
Single Rust binary install (30-second setup)
Structured JSON output for automation and CI/CD
Support for 20+ AI providers (OpenRouter, OpenAI, Anthropic, DeepSeek, Google, Ollama, and more)
Tap community registry of installable agents, skills, and MCPs
TOML configuration for per-role models, spending limits, and sandboxed execution
WebSocket server for remote programmatic AI sessions
Workflow engine chaining multiple specialist runs (octomind workflow)

At a glance

free tieropen sourceself-hostableAPI

TypeCLI tool

DeploymentHybrid

PlatformsWeb, Windows, macOS, Linux

CompanyMuvon · 2024

Integrations

OpenRouterOpenAIAnthropicDeepSeekGoogleOllama

Resources

Docs

Pricing

Pro and Max plans have discounted first-month introductory pricing ($10 first month for Pro, $50 first month for Max) before reverting to the regular monthly price. Cancelling keeps data safe for 30 days (grace period, nothing deleted). Hub model access and cloud machines are open to everyone with a free tier included, no invite or waitlist required; paid plans buy bigger caps, sizes, and slots, not access.

FreeFree

Free models only, Premium models via credits, Developer API access

$0.15/day usage allowance, free models only, ½–1 Tiny cloud machine slot, 2 concurrent sessions per machine, 2GB shared storage, 30 req/min API

Pro$20/mo

First month discounted to $10, then $20/mo (see pricing_notes).

$20/4h · $60/wk · $120/mo usage allowance, 3 cloud machine slots (Tiny, Small, Medium), 10 concurrent sessions per machine, 10GB shared storage, up to 5 custom Docker images, 180 req/min API

Max$100/mo

First month discounted to $50, then $100/mo (see pricing_notes).

$100/4h · $300/wk · $600/mo usage allowance, 12 cloud machine slots (all sizes, up to 3 Large), 20 concurrent sessions per machine, 30GB shared storage, up to 20 custom Docker images, 300 req/min API

Team$500/mo

Whole-team flat plan.

$500/mo flat for the whole team, $500/4h · $1,500/wk · $3,000/mo shared usage pool, 60 cloud machine slots (all sizes, up to 15 Large), 20 concurrent sessions per machine, 100GB shared storage, up to 50 custom Docker images, 600 req/min pooled API, unlimited team members by invite link

CreditsQuote

Metering always matches the published price table for the octomind version being run.

$5+ minimum top-up

Pro: $10 first month, then $20/mo. Max: $50 first month, then $100/mo.