forkd vs nanoclaw
nanoclaw is much bigger: 30.7k stars against 2.8k. Over the days we have tracked them forkd moved +50.8% and nanoclaw +3.6%, so forkd is growing faster right now.
nanoclaw leads on context depth and noise control and model freedom and setup ease. forkd does not take any axis by a clear margin.
Stars and commit dates come from our own daily tracking. The six axes are read off each project's documentation by our review pipeline, so they describe what a project says about itself, not what we measured in its code.
Where they stand today
Provides near-fork(2) microVM spawn and live BRANCHing via snapshot copy-on-write, enabling KVM-isolated agent fan-out with millisecond-scale latency.
- Stars
- 2.8k
- Tracked growth
- +50.8%
- Maturity
- ●●●●●
- Last commit
- 2d ago
- Language
- Rust
- License
- Apache-2.0
- Cost to run
- Free, self-hosted
Provides compact, auditable agent runtime with true OS-level container isolation so agents run in separate containers with explicit mounts instead of one large shared process.
- Stars
- 30.7k
- Tracked growth
- +3.6%
- Maturity
- ●●●●●
- Last commit
- 11h ago
- Language
- TypeScript
- License
- MIT
- Cost to run
- Your API key (Anthropic/OpenAI) or local models
Six axes, head to head
Each axis runs 0 to 5. The label under a score is what that project's own docs claim, not a category average.
| Axis | forkd | nanoclaw |
|---|---|---|
Context depth How much of your codebase it sees before it answers: the open diff, the diff plus related files, or the whole repository. | ●●●●● Diff only | ●●●●● Related files access |
Noise control How it keeps output volume down — severity thresholds, deduplication, incremental runs over new commits only. | ●●●●● None | ●●●●● Basic gating |
Customization How far it bends to your team: custom rules, prompts, style guides, per-path config. | ●●●●● CLI / REST options | ●●●●● Code-based customization |
Privacy Whether your code stays on your own infrastructure: fully local, self-hostable, or cloud API only. | ●●●●● Fully local | ●●●●● Self-hostable |
Model freedom Whether you can point it at any provider, or it is wired to one. | ●●●●● No model support | ●●●●● Bring-your-own key |
Setup ease What it takes to get a first useful run out of it. | ●●●●● Kernel + root req | ●●●●● One-script install |
Which one to pick
Pick forkd if…
Ultra-fast fan-out — pick forkd when you need hundreds of warmed, KVM-isolated microVM sandboxes and the ability to snapshot/branch a running agent mid-thought.
Pick nanoclaw if…
Privacy-first — run agents in isolated Docker containers on your own machine with per-agent mounts and the option to use local or BYO models.
- Context depth: Related files access (3/5 against 1/5)
- Noise control: Basic gating (3/5 against 1/5)
- Model freedom: Bring-your-own key (5/5 against 1/5)
- Setup ease: One-script install (5/5 against 2/5)
What people want from each one
deeplethe/forkd
Hacker News: Forkd: Fork() for AI Agent MicroVMs drew 2 points and 0 comments.
nanocoai/nanoclaw
Hacker News: Show HN: NanoClaw – “Clawdbot” in 500 lines of TS with Apple container isolation drew 533 points and 224 comments.
Questions people ask
Is forkd better than nanoclaw?
nanoclaw leads on context depth and noise control and model freedom and setup ease. forkd does not take any axis by a clear margin. forkd is worth picking when ultra-fast fan-out — pick forkd when you need hundreds of warmed, KVM-isolated microVM sandboxes and the ability to snapshot/branch a running agent mid-thought.
Which of forkd and nanoclaw keeps my code private?
forkd: Fully local (5/5). nanoclaw: Self-hostable (4/5).
What does each one cost to run?
forkd: Free, self-hosted. nanoclaw: Your API key (Anthropic/OpenAI) or local models.
Full profiles: deeplethe/forkd and nanocoai/nanoclaw. Everything else in Runtimes & sandboxes.