JEV captcha arbitrage
X post working through the economics of solving CAPTCHAs with Jev at $0.0068 per hundred versus a market paying a cent each.
This summary is based on how curated lists describe the project; its README was not available to review.
Overview
This X post works through the economics of using Jev to solve CAPTCHAs, estimating a cost of about $0.0068 per hundred solves against a marketplace that pays around a cent per solve. The yibie list presents it as a pointed illustration of what per-decision pricing does to an existing market.
It is included as a discussion of second-order effects of very cheap typed decisions rather than as a recommended use; solving CAPTCHAs at scale typically violates site terms and defeats their purpose. Only the list summary was available, so the CAPTCHA types, accuracy and method are unknown.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Choice | Which answer solves this CAPTCHA?(inferred) | — |
How it works
- Estimate per-solve cost with Jev
- Compare with marketplace payouts
- Discuss implications for that market
Strengths
- Concrete illustration of per-decision economics
Caveats
- Ethically and legally problematic use case
- No method or accuracy data available
At a glance
- Type
- Discussion
- Status
- discussion
Use cases and integrations
How curated lists describe it
X: works through the economics of solving CAPTCHAs with Jev at $0.0068 per hundred against a marketplace paying a cent each, a pointed illustration of what per-decision pricing does to an existing market.
yibie/awesome-jev
Frequently asked questions
What does the post estimate?
About $0.0068 per hundred CAPTCHA solves with Jev versus about a cent each paid by a marketplace.
Is this a recommended use?
No; it is cited to illustrate pricing effects, and CAPTCHA solving usually violates site terms.
Is accuracy reported?
Not in the list summary.
Related projects
Issue in the Orcha multi-agent orchestration project proposing to move judgment calls onto Jev to cut model cost.
ProjectdiscussionPythonBlog post that replaces an agent's classification loop with a single Jev call and reports the loop running 7x faster.
ArticlediscussionX post listing nine jobs where a typed Jev decision can replace an expensive full LLM call.
DiscussiondiscussionX account of @moritzkremb, cited in a Chinese Jev list as the source of its most specific per-call cost figure.
DiscussiondiscussionAuto-updated bilingual leaderboard of Jev ecosystem projects, tutorials and discussions, discovered daily and sorted by GitHub stars.
ProjectproductionJanus (formerly calibre): measures on your data the confidence threshold at which a small model like Jev should defer to a larger one.
ProjectbetaPython