Image JevBench v0.3.0 · individual system
GPT-6 Luna (low reasoning effort)
Closed APIBase model: undisclosedthe operator's hosted API
Image JevBench v0.3.0 composite score
12.192
Rank #26 of 55 ranked systems.
From the published full-benchmark aggregate. The composite combines intelligence, calibration, speed and cost.
Published axes
- intelligence
- 72.5
- calibration
- 76.5
- speed
- 64.6
- cost
- 25.0
Cost and speed evidence
- Cost per 1,000 decisions
- USD 0.315271
- Cost basis
- OpenRouter usage.cost receipts
- Measured latency · p50 / p95
- 4.193 s / 8.234 s
- Adjusted latency · p50 / p95
- 4.193 s / 8.234 s
- Latency adjustment
- none (hosted API, measured as is)
- Public accuracy
- 83.4%
- Sealed accuracy
- 82.1%
- Measurement note
- Official ImageJev v0.3 A300+P300 pass 1 on 7 Oct 2026 via OpenRouter chat completions (openai/gpt-6-luna, low reasoning effort, unchanged v0.2 prompt/parser, max_tokens 384). This is the chat row; the OpenAI Decisions row is measured separately. Attempt 1 aborted on a mislabelled BMP image (public items only, not scored). 548 ok, 51 unparseable answers, 1 HTTP 520. Images are sent with their real MIME type; 7 BMP images were re-encoded losslessly as PNG and 5 images over the 6 MiB request limit were downscaled to 2,048 px on the long side. Provider failures (HTTP errors, timeouts, error bodies) count wrong at cost 0; unparseable answers count wrong at their billed cost. No retries.
Published 2026-10-12. Scores and ranks can change in a later release. See the method notes.