Tag
Local tools
15 reviewed tools: 15 free or open source.
Order: free and open source first, then hosted products with a real free tier, then paid-led; catalog order inside each group, with quiet or archived repos last. Not a score, and no tool pays for its place.
15 tools
- GooseLocal-first Rust agent with MCP at its core
Free / OSS · OSS/MIT
VerdictLocal-first, MCP-native, model-agnostic. Bring your own brain.
CatchGeneral-purpose agent means you configure it for your use case. You bring the model—Goose is the runtime, not the intelligence. Documentation is still maturing.
- Milvus LiteApache-2.0 lightweight Milvus for local vector search
Free / OSS · Apache-2.0 · 465 GitHub stars
VerdictApache-2.0 local Milvus. Free for prototypes.
CatchLite is not production HA. Thinner stars than core Milvus. Distinct from full Milvus and Chroma/LanceDB peers.
- actMIT run GitHub Actions locally in Docker
Free / OSS · OSS/MIT · 72.4k GitHub stars
VerdictMIT local Actions runner. Free forever; debug the workflow before you burn a push.
CatchVs hosted GitHub Actions and Woodpecker (self-host container CI): act is a *local emulator*, not a forge runner fleet and not a replacement for GitHub-hosted minutes. Event payload fidelity is incomplete; secrets, service containers, and some actions still surprise. Needs a Docker engine.
- mirrordMIT local process as if it ran in-cluster; Teams/Operator paid
Free / OSS · OSS/MIT · 5.4k GitHub stars
VerdictMIT in-cluster local process. Free forever OSS CLI for one human; Operator/Teams and cloud-agent concurrency are the paid layer.
CatchVs Tilt (K8s *dev control plane* / live-update) and Earthly (Earthfile *builds*): mirrord is *steal the pod’s context for a local process*, not a multi-service Tiltfile and not an image builder. Teams/Operator is paid (concurrent sessions, traffic filters, queue split, DB branching, RBAC). Cloud AI Agents is Enterprise-gated. Free path is the OSS CLI for individuals.
- agentsviewMIT local-first agent session/cost archive — 20+ coding agents
Free / OSS · OSS/MIT · 6.2k GitHub stars
VerdictMIT local agent archive. Session/cost on disk; not Langfuse/Braintrust cloud.
CatchVs Langfuse / Braintrust (cloud LLM observability): agentsview is a local session/cost archive for coding-agent transcripts, not a hosted tracing/eval platform. Optional anonymous telemetry ping; you still need the agent logs on disk. Not a replacement for production OpenTelemetry of your app’s LLM calls.
- colibrìApache-2.0 pure-C zero-deps MoE streamed from disk
Free / OSS · Apache-2.0 · 42.8k GitHub stars
VerdictApache-2.0 pure-C MoE engine. Disk-streamed experts vs Ollama/vLLM; hardware-honest.
CatchVs Ollama/vLLM (local serve / high-throughput GPU serving): colibrì is research-grade MoE disk-streaming for models that do not fit, not the everyday pull-and-chat DX of Ollama or the production serving stack of vLLM. research/HW-heavy — expect fast NVMe, large model containers (hundreds of GB), and tokens/sec that track your disk hierarchy.
- HisterPrivate local search over visited pages and files
Free / OSS · AGPL-3.0 · 6.1k GitHub stars
VerdictAGPL-3.0 local privacy search. Useful personal context for agents; the network-copyleft Catch is loud.
CatchVs Gitingest and Repomix (package repositories for model context) and Agent Reach (agent-facing public-web access): Hister searches your local history/files, not a repo snapshot or the open web. AGPL-3.0 carries network copyleft if you modify Hister and offer it as a service; keep that boundary in your deployment review.
- SemIfMIT local/WebGPU semantic conditionals — OpenJev, the former pitch name
Free / OSS · OSS/MIT · 4.8k GitHub stars
VerdictMIT local semantic conditionals. Read logits on frozen open models; do not confuse it with structured output or calibrated refusal.
CatchVs Instructor and structured JSON: SemIf is a logit-read semantic gate, not schema validation or guaranteed JSON. Vs closed TypeSafe Jev: the local/open-model route is the point, but you bring compatible weights, runtime, and calibration discipline — this is not calibrated RLCD. Youth note: ~1.6k★ in ~3 days clears the Panel bar, but the surface is still young.
- SupermemoryMIT memory/context engine with a local API
Free / OSS · OSS/MIT · 31.3k GitHub stars
VerdictMIT memory/context engine. Fast local/API path; verify the paid self-host boundary before planning a hosted deployment.
CatchVs Mem0 / Letta / Graphiti: Supermemory leans into a local/API speed angle, not the same graph-orchestration shape. pricing caveat: the pricing page gates self-hosted deployments to paid tiers while the README documents a free local binary — read “self-host” as the OSS/local path, not a promise that every production deployment tier is free. You still own data policy and retrieval quality.
- Codebase Memory MCPMIT persistent code knowledge graph MCP — static binary, no LLM key
Free / OSS · OSS/MIT · 46.9k GitHub stars
VerdictMIT local code-graph MCP when agents need structural queries without a server-side model key; audit the install path and treat the ★ spike as a signal, not proof.
CatchVs CodeGraph, Serena, and Supermemory: this is a persistent local code-graph MCP from a static binary, not an AST/MCP peer alone, an LSP-symbol MCP alone, or a general memory API. Default install is curl|bash; Microsoft Defender has false-positived release binaries (Wacatac-family); star velocity is extreme (~45k, repo created 2026-02) — verify checksums/VirusTotal links before trusting the pipeline.
- MagnitudeApache-2.0 local inference desktop+CLI — consumer profiler for agents
Free / OSS · Apache-2.0 · 7.2k GitHub stars
VerdictApache local inference when agents need hardware-aware models on the desk — download magnitude.dev; ignore registry name twins and the ★ spike.
CatchDo NOT install bare npm/crates `magnitude` — name collision with unrelated packages; get the product from magnitude.dev (and the magnitudedev org), not a random registry hit. ★5,479 in about 21 weeks (~261 a week): judge it on the profiler + agent wiring, not the star count. Vs Ollama / llmfit / vLLM: Ollama is the simple local runner; llmfit sizes models to hardware; vLLM is throughput serving — Magnitude’s catch is consumer profiler + agent harness wiring in one desktop/CLI.
- LiteParseApache-2.0 local PDF/doc parser — PDFium + OCR, no cloud required
Free / OSS · Apache-2.0 · 12.9k GitHub stars
VerdictApache local parse when agents need fast PDF→Markdown on-box — not LlamaParse cloud; escalate hard scans deliberately.
CatchVs LlamaIndex (framework) and cloud LlamaParse: LiteParse is the local OSS parser binary/bindings; LlamaParse cloud is the upsell for dense tables/scans/handwriting. Hard docs still want LlamaParse or another heavy parser — do not pretend local PDFium equals production OCR SaaS.
- DwarfStar 4MIT in-process coding agent — project quants only, shared KV cache
Free / OSS · OSS/MIT · 23.9k GitHub stars
VerdictMIT in-process coding agent when you want the antirez shared-KV quants — not Ollama, not Magnitude, not llmfit.
CatchNot a generic GGUF runner. Project quants only: DeepSeek V4/V4.1, GLM 5.x, Qwen3.8. Vs Ollama / Magnitude / llmfit: Ollama is the general local model runner, Magnitude is the consumer profiler plus agent-harness wiring, llmfit only sizes models to hardware — DwarfStar 4 is the antirez in-process coding agent with a shared KV cache, not a general local runner. Still beta; pin a commit before a fleet bet.
- llama.cppMIT local LLM inference in C/C++ on almost any hardware
Free / OSS · OSS/MIT · 132k GitHub stars
VerdictThe default local LLM engine — free, runs nearly anywhere, but you own the model choice and the tuning.
CatchIt runs GGUF models, so you convert or download quantized weights, and model licences are separate — check the model card before commercial use. Lots of knobs: speed depends on backend, quant and offload settings, so expect to tune. Vs Ollama (friendlier model manager on similar ground): pick llama.cpp when you want the engine and its flags directly. Vs vLLM or SGLang: those are GPU serving engines for many concurrent users; llama.cpp is the one for laptops, edge boxes and single-user servers.
- uzuMIT on-device inference engine for Apple Silicon apps
Free / OSS · OSS/MIT · 2.4k GitHub stars
VerdictThe pick for on-device models inside an iOS or macOS app — MIT and free, but Apple-only for now and tied to its own model format.
CatchApple-first: Linux, Windows and WASM targets are still listed as in progress. It uses its own model format — you download Mirai's converted models or export one yourself with their lalamo tool, so you cannot drop in an arbitrary GGUF. The Rust quick start pulls from the git main branch. Backed by a company (Mirai). Vs Rapid-MLX (a local server for coding agents on a Mac): uzu is an embeddable engine for shipping AI inside iOS and macOS apps.
FAQ
- What is the top free pick among Local tools?
- Goose leads the Stack Gems list for Local tools. Verdict: Local-first, MCP-native, model-agnostic. Bring your own brain.
- How many Local tools does Stack Gems review?
- 15: 15 free or open source. Each review has a verdict and an honest Catch; no tool pays for its spot.
- Which Local tools can I self-host for free?
- Goose, Milvus Lite, act, mirrord, agentsview, colibrì, Hister, SemIf and 7 more.
- Which Local tools have the most GitHub stars?
- llama.cpp (132k), act (72.4k), Codebase Memory MCP (46.9k), per GitHub on 2026-10-19. Stars measure attention, not fit; read the Catch.
- What is the catch with Goose?
- General-purpose agent means you configure it for your use case.