Image JevBench v0.3.0 · individual system
Liquid d1-omni-600M (Liquid AI)
LLM decoderBase model: undisclosedlocal/GPU evaluation
Image JevBench v0.3.0 composite score
0.237
Rank #48 of 54 ranked systems.
From the published full-benchmark aggregate. The composite combines intelligence, calibration, speed and cost.
Published axes
- intelligence
- 5.7
- calibration
- 47.3
- speed
- 90.9
- cost
- 63.6
Cost and speed evidence
- Cost per 1,000 decisions
- USD 0.016367
- Cost basis
- measured GPU seconds x $1.19/GPU-hour
- Measured latency · p50 / p95
- 0.049 s / 0.087 s
- Adjusted latency · p50 / p95
- 0.247 s / 0.325 s
- Latency adjustment
- v1.4 local latency adjustment
- Public accuracy
- 31.5%
- Sealed accuracy
- 38.7%
- Measurement note
- Official ImageJev v0.3 S1200+P300 pass 1 on 10 Oct 2026 (no new draw: the same 1,500 items as every other self-hosted row) on one RTX PRO 6000 with the same reviewed runtime as the other self-hosted rows, offline after the weights were staged and hash-verified. Model: LiquidAI/d1-omni-600M at revision 02b55d70. Request: the NeoHorse-shaped single choice question through the model's own system_one() call with the item image; float16 as recommended by the model authors. 1,500/1,500 ok, no retries. With images the model cuts state and question text to 896 tokens (as trained); its context is 16,384 tokens. Same vendor as the hosted API row liquid-d1 on JevBench (text); the two rows are not linked here because the page data has no family field.
Published 2026-10-10. Scores and ranks can change in a later release. See the method notes.