Leaderboard / mini-tollbooth.onrender.com / endpoint
GET https://mini-tollbooth.onrender.com/model-picks
Scores measure reliability and honesty, not output quality.
- Answered 3 of 6 probes in the last 7 days (50%).
- Answered 3 of 6 probes in the last 30 days (50%).
- Category
- Other by llm, confidence 10%
- Latest quote
- — on — to —
- Uptime
- 50% / 7d, 50% / 30d
- HTTPS
- yes
- First listed
- 2026-11-16
- Last probe
- 5h ago
- Next probe
- 2026-11-17 16:38 UTC
Score components
Catalog listings
| Source | Scheme | Network | Asset | Listed price | Pay to | Flow | Current | Last seen |
|---|
Listing changes
| When | Event | From | To |
|---|---|---|---|
| 2h ago | disappeared | — | — |
| 2h ago | disappeared | — | — |
Paid probes
No paid probes yet. Paid probing is off by default and runs weekly per host under a strict spend cap; each paid probe links to its on-chain transaction here.
Delivery checks
Inside each paid probe we ask for something whose correct answer we know, then compare. A pass means the seller delivered what it was paid for; a fail means it charged and returned something wrong; inconclusive means we could not judge (no reference, unknown input shape, or a non-2xx response). Every row carries the request, the response hash and the reference so the verdict can be reproduced. Delivery results do not affect scores yet.
No delivery checks yet. They run inside the weekly paid probe.
Classification history
Every classification is kept; the newest row is the current category. Rules match the listing text; the model is consulted only when the rules are unsure.
| When | Category | Confidence | Method | Version | Evidence |
|---|---|---|---|---|---|
| 2h ago | Other | 10% | llm | v1 | claude-haiku-5-5: The listing has no service name, tags, description, or input fields. The path /model-picks hints at AI model recommendations but gives too little to classify reliably. |
| 2h ago | Other | 20% | rules | v1 | rule none matched "" in none |
| 15h ago | Inference | 60% | llm | v1 | claude-haiku-5-5: The endpoint ranks LLM models by value per task using quality scores and live per-token pricing, so it is model-selection intelligence about LLMs. It does not run inference itself, so inference is the closest fit, though ai-intel or pricing data could also be argued, and confidence is moderate. |
| 18h ago | Inference | 50% | rules | v1 | rule inference matched "llm" in metadata |
Probe history
| When | Kind | Status | Latency | x402 | Valid | Quoted | Pay to | Flags |
|---|---|---|---|---|---|---|---|---|
| 5h ago | free | 503 | 105 ms | — | no | — | — | server_errorstatus_not_402 |
| 11h ago | free | 503 | 121 ms | — | no | — | — | server_errorstatus_not_402 |
| 18h ago | free | 503 | 96 ms | — | no | — | — | server_errorstatus_not_402 |
| 24h ago | free | 402 | 56 ms | v2 | yes | $0.02 | 0x8BE8…96C0 | — |
| 30h ago | free | 402 | 55 ms | v2 | yes | $0.02 | 0x8BE8…96C0 | — |
| 37h ago | free | 402 | 88 ms | v2 | yes | $0.02 | 0x8BE8…96C0 | — |