Submit

Get verified.

Three ways in. The fastest is one API call — we run it, prove it, settle it on Ethereum L1. The other two are for batch listings and pre-run proofs.

Submission flow — from a developer through attestor to a published on-chain proof
Free to start
Email to get your API key. First test is free.
No activation fee, no card on signup, no subscription. Verify your email, get a bl_live_… key, post one free attested test. After that, $5 per test or load a credit pack for up to 33% off.
Option 0 · Primary
Just call the API.
Full docs →

One HTTP request posts a test. We run it on a staked attestor, Merkle-commit the transcript, submit the proof to Aligned Layer, and settle on Ethereum mainnet. You get a verify_url in the response; the listing publishes automatically once the proof verifies (~3 min).

curl -X POST https://api.benchlist.ai/v1/run \
  -H "Authorization: Bearer $BENCHLIST_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "service":   "anthropic-claude",
    "model":     "claude-opus-4-7",
    "benchmark": "mbpp",
    "runs":      3
  }'

Deducts 1 credit ($5) from your balance. Includes Ethereum L1 gas, attestor compute, and the run itself — see where your $5 goes.

Option A · No code
List a service

Tell us about your service. We run the canonical benchmarks and post the proofs. Typical TAT: 3–5 business days.
Option B · Bring your own proof
Submit a verified run

You've already run the benchmark with our runner. Paste the canonical run.json — we publish immediately after proof verification.
Option A

List a service.

We'll pick category defaults if you leave all unchecked.

Listing free during beta. Benchmarking cost recovery may apply for complex suites — see pricing.

Option B

Submit a verified run.

If you've already run a benchmark with our runner and produced a proof, paste the canonical JSON here. We publish immediately after proof verification.

Prefer the CLI? benchlist publish run.json. Docs →

Submission policy

All submissions are moderated. We reject: spam, unrelated services, listings with fabricated scores, and services that can't provide a working replay command. First offense = banned for 30 days. Repeat offense = permanent.