KNotJev : Kev

Why NotJev : Kev

Chat models are great at writing. They are awkward as product gates. NotJev : Kev turns an LLM into a typed classifier your code can threshold.

vs chat LLMs

You needChat gives youKev gives you
A labelProse you must regexA typed key + probabilities
A gate“I think so…”noul you can threshold
ConfidenceVibesConcentration of the distribution
Many judgmentsN serial promptsOne request, questions in parallel
ControlVendor lock-inYour GPU / API / laptop

vs hosted Jev / OpenJev

Hosted Jev / OpenJevNotJev : Kev
LicenseProprietary / mixedApache-2.0
DeployTheir cloud / weightsSelf-host anywhere
ModelsFixed stackBYO — Ollama, vLLM, OpenAI, …
Wire formatSystem OneSame shape (choice / score / noul)
OfflineNoMock backend for CI & demos
DXAPI keyPlayground · SDK · CLI · MCP
TrainingN/A for youNone — inference + API only

Benchmark edge

On the public OpenJev held-out suite with qwen3.5:9b, NotJev : Kev reaches 83% Banking77 (vs ~82% published Jev held-out / 80.3% JevBench), plus strong CLINC, AG News, SST-5 within-1, and toxicity noul numbers — without locking you into a closed model.

Independent project. Not affiliated with TypeSafe AI or OpenJev.