AI CLI
Vercel Labs terminal CLI for AI tasks that can use Jev as the evaluation model for its evaluate command.
This summary is based on how curated lists describe the project; its README was not available to review.
Overview
AI CLI is a Vercel Labs command-line tool for working with AI models from the terminal. According to the list summaries, its evaluate command can run Jev as the evaluation model, so judgments come back as typed probabilities rather than generated text.
The repository's top-level README only points to the package README, so details of the evaluate command, its questions and its options are not available here.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Other | Evaluation judgments run through the evaluate command (question set not documented here)(inferred) | — |
How it works
- Run the CLI's evaluate command
- Select Jev as the evaluation model
- Read typed results in the terminal
Strengths
- Brings Jev evaluation into a general AI CLI
Caveats
- Top-level README is only a pointer; details unverified
- Likely requires AI Gateway or TypeSafe credentials
At a glance
- Type
- CLI
- Language
- TypeScript
- Runtime
- CLI
- Has code
- yes
Use cases and integrations
How curated lists describe it
Developer tooling: Vercel Labs CLI that can run Jev as the evaluation model for its `evaluate` command.
yibie/awesome-jevVercel Labs terminal CLI that can run Jev as the evaluation model for its evaluate command.
hellogumbo/awesome-jev
Frequently asked questions
What does Jev do in AI CLI?
It can serve as the evaluation model for the evaluate command, per the listing.
Who maintains it?
Vercel Labs.
Where are the docs?
In packages/ai-cli/README.md in the repository.
Related projects
Node/Bun CLI for asking Jev boolean, choice and score questions over text or JSON, with direct or Vercel AI Gateway providers.
CLIbetaTypeScriptRust terminal coding agent for local models whose swarm mode can use Jev to referee parallel attempts and pick a typed winner.
LibrarybetaRustDeclarative compiler that turns state, decisions and labeled examples into a measured, frozen JSON program of typed Jev questions.
LibraryprototypePythonTool that recomputes Jev's confidence identities on outputs published in public repositories to spot answers Jev never actually produced.
ServiceresearchPythonTypeSafe's official drop-in replacement for the SDK's system_one call backed by OpenAI, Anthropic or Gemini, for comparing Jev with LLMs.
LibraryproductionPythonHack Club's lightweight LLM proxy, listed for adding a Jev forwarding endpoint behind its auth, limits and usage logging.
MCP serverproductionTypeScript