Soloco Docs
Soloco Docs
IntroductionQuickstartInstallAgent runtimesCore conceptsFeature tourFor AI agentsFAQRun AI agents locallyOpen-source AI agents: the 2026 landscapeChangelog

Run AI agents locally

What a local AI agent is, why local orchestration differs from hosted platforms, and how to run one with an existing account or your own provider key.

A local AI agent executes on your own machine and uses model access that you control — an existing account or subscription, or your own provider API key. Your workspace and orchestration state remain local by default; the selected runtime still sends the context needed for inference to its model provider. This guide explains when local beats hosted, the three ways to get there, and how to have your first local agent working on a long-running goal in about five minutes.

Why run AI agents locally?

Hosted agent platforms run your work on someone else's servers, metered by their credits. Running agents locally flips that model:

Hosted agent platformLocal AI agents
Where work happensVendor's cloudYour machine
Your files and outputsUploaded to the vendorStay on your device
Model accessPlatform credits or a provider account chosen by the platformYour Claude / ChatGPT account, or your DeepSeek API key for bundled DSH
Long-running workDepends on the vendor's runtime limitsRuns as long as your machine does
ToolingWhat the platform exposesTools exposed by your selected local runtime

Three reasons people switch to local:

  1. Local workspace control. Code, documents, orchestration state, and outputs remain on your disk by default rather than being moved into a hosted agent workspace. The selected runtime still communicates with its model provider for inference.
  2. Choose how you access the model. Reuse an existing Claude or ChatGPT account, or use your own DeepSeek API key with bundled DSH. Provider billing follows the route you choose.
  3. Real tools. A local agent can use your actual environment — your shell, your repositories, your installed CLIs — instead of a stripped-down cloud sandbox.

Three ways to run agents locally

Write your own loop. Wire an LLM API into a script that plans, executes, and retries. Maximum control, but you own scheduling, recovery, context management, and review tooling forever — and you pay per token.

Use an agent framework. Open-source frameworks (LangChain, AutoGen, CrewAI and friends) give you building blocks for multi-agent systems in code. Great when you are building an agent product; heavy when you just want agents to do your work, and most still bill through API keys.

Orchestrate local agent runtimes. Claude Code and Codex CLI are capable local agents that can use an existing account or subscription. Soloco also bundles a compatible DeepSeek Harness for users who bring their own DeepSeek API key. Soloco takes this third path: it does not run its own hosted agents — it coordinates the selected runtime on your machine, turning a goal into an executable, auditable, resumable task tree. The runtime brings execution; Soloco brings the organization.

Set up a local agent in three steps

The short version — the Quickstart has the full walkthrough:

  1. Choose a supported runtime: install and sign in to Claude Code or Codex CLI, or use the DSH runtime bundled with Soloco on macOS or Linux and enter your own DeepSeek API key. Compare the routes in Agent runtimes.
  2. Install Soloco: npm i -g @soloco/client@latest (Node.js ≥ 22.5 on macOS, Windows, or Linux — see Install).
  3. Run soloco, open the local UI, and describe a goal. Soloco plans it into tasks, dispatches them to your runtime, verifies results, and keeps the goal moving — even while you are away from the keyboard. See Core concepts for how the task tree, standing missions, and heartbeats fit together.

Frequently asked questions

Can a local agent keep working when I close my laptop? Not while it sleeps — agents run while your machine is awake. State is persisted locally, so when the machine wakes, a goal resumes where it left off instead of restarting.

Is this open source? Soloco's V2 core is closed-source, while registration is open. The runtimes it orchestrates (and many frameworks in the table above) have their own licenses. Pricing is the source of truth for cost.

What data leaves my machine? Your workspace, orchestration state, and outputs remain local by default. The selected runtime sends the context needed for inference to its model provider: Claude Code and Codex CLI use their configured accounts, while bundled DSH uses the DeepSeek API key you provide.

For a machine-readable summary of these facts (useful if an AI assistant is evaluating Soloco for you), see /agent or /llms.txt. For what Soloco is not, read Introduction.

Table of Contents

Why run AI agents locally?Three ways to run agents locallySet up a local agent in three stepsFrequently asked questions