The best open-source tools for coding agents in 2026

The best open-source tools for coding agents right now are ccusage, CC Safety Net and Claude Code Router, out of 10 we have reviewed. Each is ranked by our rating, with its price, our verdict and who it suits.

By Sheela · Updated Oct 31, 2026 · How we review

How we ranked: by our rating out of 5, from reviews built on each product's own pricing, documentation and policies plus independent benchmarks where they exist, each checked on the date shown in the review. We take no affiliate commission.

1. 1
   

   ccusage

   ccusage is the meter every heavy agent user should run, provided they read its dollar column as what the work would have cost on the API, not what they paid.

   Best for: Anyone who wants to track or compare token use across agent sessions.

   Price: Free and open source (MIT).

   Read our ccusage review
   4.4/5
2. 2
   

   CC Safety Net

   The seatbelt for the damage a sandbox allows on purpose: force-pushes, hard resets and rm -rf inside your own project.

   Best for: Anyone letting a coding agent run shell commands without approving each one.

   Price: Free and open source under MIT.

   Read our CC Safety Net review
   4.0/5
3. 3
   

   Claude Code Router

   A capable local switchboard for running coding agents on cheaper models, as long as you accept that Anthropic will not support the result and the edits will be worse.

   Best for: Developers who want to run Claude Code or other agents on cheaper or local models.

   Price: Free and open source (MIT). You pay each model provider you route to.

   Read our Claude Code Router review
   4.0/5
4. 4
   

   claude-real-video (crv)

   The sensible way to let Claude reason about a video today: local, cheap on tokens and honest about timestamps, though it shows the model stills and words rather than motion.

   Best for: Claude Code and other agent users who need a model to answer questions about videos without uploading them.

   Price: Free and open source (MIT). Optional crv Pro is a one-time $29; a separate Codex build is a one-time $9.90.

   Read our claude-real-video (crv) review
   4.0/5
5. 5
   

   notebooklm-py

   NotebookLM's full feature set in an agent's hands, held together by undocumented Google endpoints and a login you should not give it lightly.

   Best for: Researchers who want Claude Code to build and query NotebookLM notebooks.

   Price: Free and open source (MIT). NotebookLM limits depend on your Google account tier.

   Read our notebooklm-py review
   3.8/5
6. 6
   

   claude-mem

   Real cross-session memory for Claude Code, wrapped in an installer that steers you towards a hosted subscription and a compression bill you have to watch.

   Best for: Developers on long-running projects who want the agent to remember past sessions.

   Price: Open source (Apache-2.0). Compression runs on your Anthropic plan or your own key. Hosted CMEM plans: Pro $30/month, Heavy $70/month, Max $100/month with overages; TeamBrain pricing not announced.

   Read our claude-mem review
   3.7/5
7. 7
   

   context-mode

   Worth trying for long sessions that lose the thread after compaction; nobody has yet shown the free plugin lowers your bill.

   Best for: Claude Code users whose long sessions lose track of their work after compaction.

   Price: Plugin free (Elastic License 2.0). Hosted gateway: 1,000 free requests once; Starter $20/month (10,000 requests); Pro $49/month (50,000 requests); you pay your model provider directly.

   Read our context-mode review
   3.6/5
8. 8
   

   mindwalk

   Seeing where an agent wandered in your repository takes seconds with mindwalk, but it is a young tool whose development has gone quiet since August.

   Best for: Developers and tech leads who review coding-agent work and want to see an agent's footprint before reading its diff.

   Price: Free and open source (MIT). Session evaluation uses your existing Claude or Codex plan.

   Read our mindwalk review
   3.6/5
9. 9
   

   rtk (Rust Token Killer)

   It makes shell output shorter, which is not the same as making Claude Code cheaper, and both independent cost tests found it was not.

   Best for: Developers with very noisy shell output who will measure the effect themselves.

   Price: Free (Apache-2.0 CLI plus BUSL-1.1 engine; Apache-only rtk-oss available). Commercial resale needs a licence. RTK Pro team edition is paid, with a 7-day trial.

   Read our rtk (Rust Token Killer) review
   3.4/5
10. 10
    

    Agent Reach

    Agent Reach is a clever convenience for personal research on Chinese and social platforms, and a liability anywhere an employer, a platform's terms or your main account matter.

    Best for: Individual researchers and developers who need their agent to read Chinese platforms and accept the account and terms-of-service risk on a secondary account.

    Price: Free under MIT. Some channels need your own keys: YouTube transcription uses a Groq or OpenAI key, podcast transcription a Groq key. The newest CloakBrowser binary needs a paid CloakHQ subscription.

    Read our Agent Reach review
    3.2/5