Tokens for Good vs Folding@home
If you ever ran Folding@home, you already understand Tokens for Good. Both turn capacity you are not using into a shared public good; one used idle GPUs, the other uses idle AI.
The shared idea: give what you are not using
Folding@home is volunteer computing. Thousands of people let their idle CPUs and GPUs crunch through pieces of a huge scientific problem, like simulating how proteins fold, while their machines would otherwise sit doing nothing. Tokens for Good borrows that same spirit: it is Folding@home for the AI era. Instead of spare processor cycles, you give the spare AI capacity in your Claude subscription.
What each one actually crunches
The work is different because the resource is different:
- Folding@home: your hardware runs raw math, simulating molecular behavior for disease research.
- Tokens for Good: your AI agent does knowledge work, researching a nonprofit's real impact against a fixed methodology and writing a cited report.
One contributes computation; the other contributes reasoning and research.
The output: science vs a public directory
Folding@home feeds results back to researchers studying disease. Tokens for Good feeds a public directory of vetted nonprofits. Every organization is researched twice by independent contributors, validated for unsupported claims, consolidated into one scorecard, scored deterministically, and finalized by a human reviewer. The result is something a giver can actually use to decide where to put their money or time. See the full pipeline.
Setup and cost compared
Folding@home asks you to install a client and leave your computer on. Tokens for Good is similar in feel but lighter: run npx tokens-for-good init or add the remote MCP, then optionally schedule runs. Scheduled work executes on Anthropic cloud, so your own machine can be off entirely. Neither costs you extra; Folding@home uses electricity you are already paying for, and Tokens for Good uses the Claude plan you already have.
Which should you run?
They are not rivals; they help different causes. If you have a powerful GPU sitting idle and want to back scientific research, Folding@home is great. If you pay for Claude and rarely max out your limits, Tokens for Good turns that headroom into AI for good. Many developers will find Tokens for Good the more natural fit simply because their AI subscription is the thing going unused.
Frequently asked questions
Is Tokens for Good run by Folding@home?
Does Tokens for Good use my CPU or GPU like Folding@home?
Is either one cryptocurrency mining?
Can I run both at the same time?
Volunteer your AI, not just your GPU
Tokens for Good applies the volunteer-computing idea to the AI capacity you already pay for.
See how it works