Pricing
Free. All of it.
The score, claiming your listing, the badge, the improve dashboard and the API are free. There is no paid tier.
What is included
Scoring and a public listing
Every listed agent is scored by the published rubric and shown on the leaderboard with its evidence class. Scores are re-assessed against each agent's public materials; changes of more than 5 points are checked by a person before they publish. We are not adding new agents right now. See the scoring methodology.
Claiming and the improve dashboard
Claim your listing to see the dimension-level breakdown, the assessor's rationale, and a ranked list of the changes most likely to move your score.
The embeddable badge
A score badge for your README or site. It states the score's evidence class in the badge itself, so it cannot read as more than it is.
API access
Create a key from your dashboard and read the catalog programmatically. Keys are rate-limited to 100 requests per minute. See the API docs.
How this is paid for
No paid tiers, no sponsors, and no vendor money. No agent operator pays BenchLytix for a score, a rank, a placement or a badge, so none of those can be bought.
BenchLytix is built by the same team that builds Metrx, and Metrx production telemetry is the planned source of a runtime tier, which is not live: no score on BenchLytix uses it. We state that relationship wherever that data would appear rather than leaving it to be discovered. The methodology, the evidence class of every score and the assessor models used are published so that the output can be checked rather than trusted.
If any of this changes, it changes on this page first.
What comes next, also free
A tested open-source-versus-SaaS scorecard: a published task suite run against both tools, with the traces, so a buyer can see whether the free tool covers the job before cancelling the paid one. The first category is n8n versus Zapier. Nothing on that page will be sponsored either.
Frequently asked questions
- Does anyone pay for rank, placement or a badge?
- No. There is no paid tier, and placement is not for sale. Scores are computed strictly from the published rubric (see the scoring methodology), and the same pipeline scores every agent.
- Is the badge free?
- Yes. The badge shows the score and the score's evidence class (a desk estimate says so; only a runtime-measured score reads Verified). It is earned, not bought, and it does not change how the score is calculated.
- Will there ever be paid features?
- None are offered today. If we introduce one, it will be announced on this page before anything is charged, and it will never affect a score, a rank or a placement.
- What is the relationship with Metrx?
- BenchLytix is built by the same team that builds Metrx (metrxbot.com). Metrx production telemetry is the planned source of a runtime tier, which is not live: no score on BenchLytix uses it. The relationship is disclosed on every surface that would use that data; see about.
More on independence in about and the scoring methodology.