Hopper
This detail uses the public, hash-checked v1.4.1 aggregate. Scores and ranks can change when a new release is published; the page preview remains name-only.
Published result
- Rank
- 3
- JevBench Score
- 59.4
- Sealed-set accuracy
- 34.1%
- Cost
- $0.024 per 1,000 decisions (estimated)
Published axes
- intelligence
- 48.0
- calibration
- 79.1
- speed
- 86.8
- cost
- 58.7
Availability and evidence
- Openness
- Code and weights marked open in the published row
- License note
- Component-specific terms recorded in RESULT.md; submitted adapter release and Qwen base retain their respective terms
- Cost evidence
- estimated; the board’s row disclosure contains the published basis.
- Endpoint condition
- our GPU (lium.io RTX A6000 48 GB), local loopback HTTP; serial
- Published source
- https://huggingface.co/HopitAI/hopper
Read the full board and published method. The overall score is a composite, not raw accuracy.