Tools with a free plan or free software, ranked by how often Hacker News commenters recommended them over the last three years in our sample. The ranking comes from HN comments; the plan details come from each vendor's official page, like every listing on this site.
30 listings · facts checked on official pages, latest 2026-11-02. Plans change; confirm on the vendor's site.
How we ranked them
We searched three years of Hacker News comments (since November 2023) through the public HN Algolia search API, with five recommendation-style queries per tool, and counted distinct comments where the tool name sits next to a phrase like “I recommend”, “I use”, “switched to” or “free tier”. The counts come from a sample of search results (at most about 500 per tool), so they are relative, not totals. Some matches are weak, and only tools from our candidate list that have a verified free plan or are free software on this site are ranked. Counts were taken on 2026-11-01.
The open-source OpenCode agent for your terminal, IDE or desktop, with limited-time free models on OpenCode Console (such as Beam Free, GLM-5.3-Flash Free, DeepSeek V4.1 Flash Free, MiMo-V2.6-Flash Free, Nemotron 3.5 Lightning Free and Big Pickle) or any provider's models through your own key.
How to get it
Install with the script, npm, bun or brew from opencode.ai.
The catch
Free models are offered for a limited time; using OpenCode's pay-as-you-go models means adding billing details.
A private WireGuard-based network between your devices and servers: unlimited user devices for up to 6 users, 3 ACL groups, 50 tagged resources and 1,000 minutes a month for ephemeral resources.
How to get it
Click "Get started" under Personal on tailscale.com/pricing.
What it requires
Non-commercial use; the FAQ says the Personal plan is for individuals using Tailscale at home.
The catch
The Personal plan is only for non-commercial use; business use starts on Standard ($8 per user a month).
Gemini app with Auto model selection (most prompts use Flash-Lite; ones that need deeper reasoning may use Flash or Pro), image generation and editing, Deep Research, Gemini Live, Canvas, Gems, limited Google Flow access and 15 GB of storage.
How to get it
Sign in at gemini.google with a Google Account.
What it requires
A Google Account.
The catch
Most free prompts run on Flash-Lite; with smart model selection off, every reply uses Flash-Lite. Exact Free limits are not published.
Unlimited storage of logins, notes, cards and identities, access on any device and a password generator; a free organization lets you share with one other person.
How to get it
Sign up for a free account at bitwarden.com.
What it requires
Not stated in the plans article.
The catch
Advanced 2FA options, the Bitwarden authenticator (TOTP), encrypted file attachments and emergency access need Premium.
Entry-level models (auto model selection), 2,000 code completions a month, limited chat and agent mode, and Copilot CLI.
How to get it
Get started at github.com/features/copilot.
What it requires
A GitHub account.
The catch
Chat and agent use draw on a small GitHub AI Credits allowance. More use needs Pro ($10/month); frontier models need Pro+ ($39/month) or Max ($100/month).
Online editor for the Typst typesetting language: create, share and collaborate on projects, community packages and templates, conversion from LaTeX, Word and Markdown, 200MB storage. The Typst compiler itself is open source and free to run locally.
How to get it
Sign up under Free on typst.app/pricing, or install the open-source CLI.
The catch
Free projects get 200MB storage and 100 files each. Comments, private packages, GitHub/GitLab sync and 2GB storage need Pro at $7.99 a month.
Practically unlimited basic searches, 3 Pro Searches a day, 1 Research query a month, limited basic file uploads and search history. Perplexity picks the model for each query.
How to get it
Sign up at perplexity.ai.
The catch
No choice of advanced models and no image generation on Free.
Light quota to code with agents, limited model availability, unlimited inline edits and unlimited Tab completions.
How to get it
Choose Download under Free on devin.ai/pricing.
The catch
windsurf.com/pricing now redirects to devin.ai/pricing, where the editor is sold as Devin Desktop. The quota size isn't published, and frontier models need Pro ($20/month).
Hub hosting for models, datasets and Spaces, with Spaces on free CPU Basic (2 vCPU, 16 GB) and ZeroGPU hardware: 5 minutes of ZeroGPU time a day on a free account, and up to 2 hosted ZeroGPU Spaces.
How to get it
Sign up at huggingface.co.
What it requires
A free Hugging Face account.
The catch
Free accounts get no included Inference Providers credits; you buy credits or upgrade to PRO ($9/month with $2 of credits a month).
A self-hosted ChatGPT-style interface for local models through Ollama or any OpenAI-compatible API, with RAG, tools and multi-user support, able to run fully offline.
How to get it
Install with Docker or pip from docs.openwebui.com.
What it requires
Your own server, plus local models or an AI provider API key.
The catch
Not standard open source: the Open WebUI License requires keeping the Open WebUI branding in deployments with more than 50 users. You pay for any hosted models you connect.
Git repositories, issues, pull requests, packages and Codeberg Pages hosting for free/libre software projects, run by the non-profit Codeberg e.V.
How to get it
Create an account on codeberg.org.
What it requires
Projects should use a free/open-source license; private content is limited to 100 MB for contributors to free software.
The catch
Not for private or proprietary work: commercial projects needing private repos are pointed to self-hosting Forgejo. Larger projects are reviewed and can be denied service.
Open-source, self-hosted PaaS that deploys apps, databases and one-click services to your own servers, with unlimited servers, deployments, teams and members.
How to get it
Click Start self-hosting on coolify.io/pricing and install it on your server.
What it requires
Your own server (a VPS, Raspberry Pi or old laptop).
The catch
You run, back up and update Coolify yourself and pay for your own servers. Managed Coolify Cloud is $5 a month for 2 servers plus $3 per extra server.
We searched Hacker News comments from the last three years through the public HN Algolia search API, running five recommendation-style queries per tool (such as "I use", "switched to" and "free tier") and reading up to 100 results each. A comment counts once for a tool if the tool name appears close to a phrase like recommend, I use, switched to, happy with or free tier.
Are these total mention counts?
No. The numbers are distinct matching comments within a sample of search results, capped at about 500 per tool, so they are a relative signal for ordering, not a count of every HN comment. Some matches are weak, and tools outside our hand-picked candidate list could not appear.
Why are some popular HN favorites missing?
Only tools listed on Free Tier Finder with a verified free plan or free software are ranked. Paid-only favorites such as Kagi, Fastmail and Hetzner were left out, as were open-source tools not listed here yet. Claude Code was excluded because it is not part of the Claude free plan.