k-dense-byok vs MiroThinker
MiroThinker is much bigger: 8.4k stars against 1.2k. Over the days we have tracked them k-dense-byok moved +30.6% and MiroThinker +1.3%, so k-dense-byok is growing faster right now.
k-dense-byok leads on context depth and model freedom. MiroThinker 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
A local-first, bring-your-own-keys AI research assistant that runs on your computer and bundles domain-specific scientific skills, workflows, and a living lab notebook.
- Stars
- 1.2k
- Tracked growth
- +30.6%
- Maturity
- ●●●●●
- Last commit
- 2d ago
- Language
- TypeScript
- License
- MIT
- Cost to run
- Your API key (pay-per-use) or free local models
Provides open-source deep research agents with extremely long (256K) context windows and very high tool-call budgets, enabling long‑horizon, verifiable multi-step research workflows.
- Stars
- 8.4k
- Tracked growth
- +1.3%
- Maturity
- ●●●●●
- Last commit
- 69d ago
- Language
- Python
- License
- Apache-2.0
- Cost to run
- Free; self-host models or use HuggingFace
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 | k-dense-byok | MiroThinker |
|---|---|---|
Context depth How much of your codebase it sees before it answers: the open diff, the diff plus related files, or the whole repository. | ●●●●● Whole-project access | ●●●●● Diff only |
Noise control How it keeps output volume down — severity thresholds, deduplication, incremental runs over new commits only. | ●●●●● Clarifying prompts & review | ●●●●● Trace collection |
Customization How far it bends to your team: custom rules, prompts, style guides, per-path config. | ●●●●● Configurable skills & presets | ●●●●● Configurable workflows |
Privacy Whether your code stays on your own infrastructure: fully local, self-hostable, or cloud API only. | ●●●●● Fully local optional | ●●●●● Fully local |
Model freedom Whether you can point it at any provider, or it is wired to one. | ●●●●● BYOK + local models | ●●●●● HuggingFace weights |
Setup ease What it takes to get a first useful run out of it. | ●●●●● Quick local start | ●●●●● Online demo |
Which one to pick
Pick k-dense-byok if…
Privacy-first — run a full scientific assistant locally or use your own OpenRouter key while keeping projects and data on your machine, with many prebuilt scientific skills and workflows.
- Context depth: Whole-project access (5/5 against 1/5)
- Model freedom: BYOK + local models (5/5 against 3/5)
Pick MiroThinker if…
Long‑context research — choose MiroThinker when you need an open-source agent with 256K context and high tool‑call capacity for deep, verifiable multi‑step research tasks.
What people want from each one
K-Dense-AI/k-dense-byok
Questions people ask
Is k-dense-byok better than MiroThinker?
k-dense-byok leads on context depth and model freedom. MiroThinker does not take any axis by a clear margin. k-dense-byok is worth picking when privacy-first — run a full scientific assistant locally or use your own OpenRouter key while keeping projects and data on your machine, with many prebuilt scientific skills and workflows.
Which of k-dense-byok and MiroThinker keeps my code private?
k-dense-byok: Fully local optional (5/5). MiroThinker: Fully local (5/5).
What does each one cost to run?
k-dense-byok: Your API key (pay-per-use) or free local models. MiroThinker: Free; self-host models or use HuggingFace.
Full profiles: K-Dense-AI/k-dense-byok and MiroMindAI/MiroThinker. Everything else in Research agents.