Safety Findings

Safety findings preserve a source’s model, benchmark, hazard and severity context. Findings are scoped observations, not a universal assurance that a model is safe or unsafe.

Use source and benchmark filters to examine comparable evidence. Model-name, hazard and severity searches are case-insensitive substrings. Keep source links and observation dates visible when displaying findings.

Official AILuminate results use source ailuminateResults. Their ordinal labels appear in grade; score remains null. mlCommons is accepted as a legacy read filter for this source family, preserving original source IDs and history in returned findings. The existing Benchmarks source page redirects /benchmarks/mlCommons to /benchmarks/ailuminateResults.

AI Incident Database cases use the separate typed benchmark context API. Reported real-world incidents are not model safety-test scores and must not be averaged into the AILuminate grade scale or model leaderboards.

On this page

List safety findings

GET /api/v1/safety/findings

Access: public read; no TensorTrace API key.

Parameters

NameLocationRequiredType / limitsMeaning
sourcequeryNostringRegistered data-source ID, such as openRouter. This identifies the collecting source, not necessarily the model developer.
modelqueryNostringCase-insensitive substring of the source-reported model name.
benchmarkqueryNostringCase-insensitive substring of the benchmark name. It does not merge different benchmark versions or conditions.
hazardqueryNostringCase-insensitive substring of the reported safety hazard.
severityqueryNostringCase-insensitive substring of the reported safety severity.
limitqueryNointeger Default: 50; Minimum: 1; Maximum: 200Maximum number of returned rows. The documented default and range apply; numeric pagination limits are clamped unless stated otherwise.
offsetqueryNointeger Default: 0; Minimum: 0Zero-based number of matching rows to skip. Default 0; negative values are clamped to 0.

Request example

curl 'https://tensortrace.net/api/v1/safety/findings?limit=1'

Responses

StatusMeaningBody type
200Source-backed JSON responseSafetyFindingsResponse
400Invalid queryNo declared JSON body
404Unknown model/sourceNo declared JSON body
500Database unavailableNo declared JSON body
503Service temporarily unavailableNo declared JSON body

Response example · 200

Illustrative values; schema-valid and not a live catalog measurement.

{
    "data": [
        {
            "source_id": "openRouter",
            "source_record_id": "google/gemini-2.5-flash",
            "benchmark_id": "example",
            "benchmark_name": "example",
            "hazard_category": "example",
            "metric": "example",
            "unit": "input_tokens_per_million",
            "updated_at": "2026-10-05T00:00:00Z",
            "canonical_model_id": "google/gemini-2.5-flash"
        }
    ],
    "total_count": 1
}
FieldTypeRequiredMeaning / constraints
dataarray of SafetyFindingDtoYes
total_countintegerYes

Models

Loading
EVIDENCE / COMPARISON

Benchmarks

Explore every benchmark

Loading

These charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.

Install TensorTrace