Stack Gems
Menu
WorkbenchBestNewSkillsLogAboutSubmit

Best of · AI Agents

Best AI agent tooling (2026)

Stack Gems reviews 4 AI agent tooling: 4 free or open source. These are the top 4, each with its verdict and the Catch we would want to know before installing it.

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.

4 reviewed · 4 shown in full

  1. 1.llmfitMIT CLI — which models fit your hardware

    Free / OSS · OSS/MIT · 38.1k GitHub stars

    VerdictMIT hardware-fit CLI. Pick models that fit; run them elsewhere.

    CatchVs Ollama / llama.cpp / vLLM (runtimes): llmfit is the hardware fit finder, not the model runtime. Catalogs and VRAM heuristics drift; always verify against your real device and quantization. Does not download or serve models for you.

  2. 2.Knowledge Work PluginsApache-2.0 knowledge-work plugins (Anthropic)

    Free / OSS · Apache-2.0 · 30.3k GitHub stars

    VerdictApache-2.0 knowledge-work plugins. Free pack; Cowork/paid host is the intended runtime.

    CatchPlugins are free; primary intent is Claude Cowork (paid host). Vs generic Claude Code skills packs: this tree is aimed at knowledge-work / Cowork surfaces, not a drop-in replacement for every coding-agent skill. You still need a compatible Anthropic host product and whatever model plan that host requires.

  3. 3.Weave RouterELv2 on-box embedder proxy that routes each prompt to the right model

    Free / OSS · Open-Core · 5.6k GitHub stars

    VerdictELv2 source-available model router. Strong cost lever for agent fleets; no managed offering without a license review.

    CatchElastic License 2.0 is source-available, not OSI open-source — you may not offer Weave Router as a hosted/managed service under ELv2 without permission. Vs LiteLLM: LiteLLM is the broader multi-provider gateway/SDK; Weave Router's wedge is per-prompt model selection via on-box embeddings. Routing quality still depends on your traffic mix and evaluation.

  4. 4.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.

FAQ

What is the top free pick among AI agent tooling?
llmfit leads the Stack Gems list for AI agent tooling. Verdict: MIT hardware-fit CLI. Pick models that fit; run them elsewhere.
How many AI agent tooling does Stack Gems review?
4: 4 free or open source. Each review has a verdict and an honest Catch; no tool pays for its spot.
Which AI agent tooling can I self-host for free?
llmfit, Knowledge Work Plugins and LiteParse.
Which AI agent tooling have the most GitHub stars?
llmfit (38.1k), Knowledge Work Plugins (30.3k), LiteParse (12.9k), per GitHub on 2026-10-19. Stars measure attention, not fit; read the Catch.
What is the catch with llmfit?
Vs Ollama / llama.cpp / vLLM (runtimes): llmfit is the hardware fit finder, not the model runtime.

Alternatives

Tags in this list

Related lists

JSON: /api/gems?domain=ai-agents&subcategory=tooling · MCP: best_for on /api/mcp