goose vs NanoClaw
Price, platforms and features side by side. The table says where each price was read and when.
Where they differ
- goose is free. NanoClaw is free.
- NanoClaw is self-hostable, according to its own site.
- NanoClaw also covers chat assistant and templates.
In commonFree planmacOSWindowsLinuxOpen sourceWorkflow automationAI agents
At a glance
| goose | NanoClaw | |
|---|---|---|
| Pricing | Free | Free |
| Starts at | Free | Free |
| Free plan | Goose is free and open source under the Apache 2.0 license, with no paid plan. Model usage is billed by the provider you connect, unless you run a local model or use a provider's free tier. | The whole product is free and open source under the MIT license. You supply the computer it runs on and your own AI provider account, which the provider bills separately. |
| Free trial | None listed | None listed |
| macOS | ||
| Windows | ||
| Linux | ||
| Made by | Agentic AI Foundation (AAIF) | NanoCo |
| Price read from | goose-docs.ai | nanoclaw.dev |
| Checked | Oct 9, 2026 | Oct 6, 2026 |
Who each one suits
goose
- Developers who want a local agent for coding and terminal work
- People who want to reuse a Claude, ChatGPT or Gemini subscription they already pay for
- Teams that share repeatable workflows as recipes
NanoClaw
- People who want a personal agent running on hardware they control
- Developers who want a small codebase they can read and fork
- Giving an agent shell access inside a container instead of on the host
Worth knowing
goose
- goose is free, but hosted models are billed by the provider you connect, and the docs note that not every provider has a free tier.
- It runs on your own machine: there are desktop apps for macOS, Linux and Windows and a CLI, and the site lists no hosted cloud version.
- The docs say goose relies heavily on tool calling and currently works best with Claude 4 models.
- The project left Block for the Agentic AI Foundation in April 2026; its repository moved from block/goose to aaif-goose/goose.
NanoClaw
- There is no hosted version: you install it yourself with git and a terminal, and the computer must stay on for the agent to respond.
- The software is free, but hosted model usage is billed by the provider you connect, through a subscription or an API key.
- Docker is the only supported container runtime, and Windows is supported only through WSL2.
- The site lists no graphical app of its own: you talk to it in the terminal or through the messaging channels you connect.
Key features
goose
- Desktop app, CLI and API
- The same agent runs in a native desktop app, a terminal CLI and an API for embedding. The desktop app and CLI share provider, model and extension settings.
- MCP extensions
- More than 70 documented extensions connect goose to databases, APIs, browsers, GitHub, Google Drive and other tools through the Model Context Protocol, and you can build your own.
- Providers and existing subscriptions
- Works with 15+ providers, including Anthropic, OpenAI, Google, Ollama, OpenRouter, Azure and Bedrock, or with a Claude, ChatGPT or Gemini subscription you already have, through ACP.
- Built-in local inference
- The desktop app can download a GGUF model from Hugging Face and run it in-process with llama.cpp, with no API key and no separate model server.
- Recipes
- Workflows are saved as portable YAML recipes holding instructions, extensions, parameters and subrecipes, then shared with a team or run in CI.
- Subagents and safety controls
- Subagents handle parallel tasks outside the main conversation. Safeguards include prompt injection detection, tool permission controls, a sandbox mode and an adversary reviewer.
NanoClaw
- Messaging channels
- Talk to agents in the terminal or connect WhatsApp, Telegram, Slack, Discord, Signal, iMessage, Microsoft Teams and other channels, each installed on demand with an /add- skill.
- Container isolation
- Every active session runs in its own unprivileged Docker container and sees only the files you mount, so shell commands run inside the container rather than on your host.
- Credential vault
- By default agents never hold raw API keys: outbound requests pass through OneCLI's Agent Vault, which injects credentials at request time and applies per-agent policies and rate limits.
- Scheduled tasks
- Ask an agent in chat for a reminder or a recurring job and it creates a task on a cron-style schedule. Tasks can also be listed, paused and cancelled with the ncl admin CLI.
- Choice of model provider
- Claude is the default, Codex can be chosen during setup, and OpenCode or local Ollama models can be added later. Each agent group can use a different provider.
- Templates and skills
- A template stamps out a prebuilt agent with a persona, skills, MCP servers and paused recurring tasks, while skills add channels, providers or tools to your own fork.
Plans
goose
goose (Apache 2.0)Free
The full desktop app, CLI and API. You bring a model provider, an existing AI subscription or a local model.
NanoClaw
NanoClaw (MIT licensed)Free
The complete self-hosted agent, with channels, providers and tools added as skills. Model usage is billed by the provider you connect, unless you run local Ollama models.
