For AI agents: the complete documentation index is available at https://viteval.dev/llms.txt, the full documentation bundle is available at https://viteval.dev/llms-full.txt, and this page is available as Markdown at https://viteval.dev/cli/run.md.
V1 release candidate available.Migration guide →
GitHub

run

Run tasks and scorers with viteval.config.ts. Vitest is the internal execution engine, not a public configuration or CLI passthrough.

Usage

viteval run [filters...] [options]
viteval [filters...] [options]

Both forms run once. Use viteval run --watch for continuous evaluation.

Arguments

ArgumentDescription
filters...File path filters passed to Vitest, not glob patterns

Options

Viteval owns the option schema. Per-evaluation options override CLI flags, which override project defaults. Unknown options fail rather than forwarding to Vitest.

OptionPurpose
--watchRerun evaluations when files change
--name <pattern>Filter evaluation names
--retries <count>Retry failed cases
--timeout <ms>Set the per-case timeout
--concurrency <count>Limit concurrent cases
--root <path>Select the project root
--config <path>Select a Viteval configuration file
--coverageCollect code coverage

Display the complete option reference:

viteval run --help

Examples

viteval run
viteval run chat.eval.ts summarization.eval.ts
viteval --watch
viteval run --timeout 60000 --retries 2
viteval list
viteval list --json

A failed case or persistence failure produces a nonzero exit code. Select eval.results: ['json'] to save evaluation details. Listing imports modules but does not execute generators, tasks, or scorers. See run in CI and inspect results.

Compiled executable

The repository can compile Viteval into an executable using maltty's Bun bundler. See the build instructions.

The executable handles help, version, and initialization directly. Evaluation, dataset, and UI commands use the project's installed viteval package through Node >=26. Install viteval in the project before using those commands.

For all project commands, use viteval --help.