Jev reviews PRs 1.93x faster at $0.0014 per run vs GPT-5.6 Luna
aniketmaurya · x · 2026-09-20
An open-source benchmark from CelestoAI compares Jev against GPT-5.6 Luna on GitHub pull request review:
- Jev ran 1.93x faster
- Cost just $0.0014 per review
- Full code available in the pr-review-jev directory of CelestoAI/celesto
Review Lab lets you paste a public GitHub PR and watch an agent prep the test environment, run checks, and compare two reviewer models side by side. It supports local Celesto OSS or Celesto Cloud; Jev defaults to Vercel Gateway via AIGATEWAYAPIKEY, and API keys never enter the sandboxes. Requires Python 3.12+, uv, and one sandbox backend (plus Node 22+ for the Vercel path).
More from coding & agent
- Accorduon turns a foldable iPhone into an accordion — the hinge is the bellows — rounak · 2026-09-20
- Agents no longer need setup: hand them a bare machine and they fetch everything themselves — vivekhaldar · 2026-09-20
- DiffusionGemma 26B-A4B turned into a local System One fast-decision model via vLLM patch — solyarisoftware · 2026-09-20
- Devin's New SWE 2 Model Goes Free and Unlimited, Claims Kimi K3 Post-Training — silasalberti · 2026-09-20
- Dev uses Claude to label 200 unsupervised visual clusters in a painting-retrieval app — jamievurnilla · 2026-09-20
- Open-source semgrep finds code and logs by meaning, not regex, with cross-language matching — udmrzn · 2026-09-20