k6 2.0 Unleashes AI-Driven Performance Testing with New Agentic Workflows

Breaking: k6 2.0 Now Generally Available with AI-Assisted Testing Capabilities

Grafana Labs today announced the general availability of k6 2.0, a major update to the open-source performance testing tool that introduces AI-assisted workflows, enhanced browser testing, and a new Assertions API. The release aims to help teams keep pace with faster development cycles powered by artificial intelligence.

“With AI assistants generating code at unprecedented speed, testing must evolve to remain frictionless and interpretable by both humans and agents,” said Sarah Liu, VP of Engineering at Grafana Labs. “k6 2.0 is built from the ground up to let teams create tests faster, express expectations more clearly, and scale validation seamlessly from local development to production-like environments.”

Key Features in k6 2.0

The update introduces four new k6 x commands designed to integrate with AI coding tools like Claude Code, Codex, and Cursor. The k6 x agent command bootstraps agentic testing workflows, setting up configuration and skills an agent needs to write correct, idiomatic tests. k6 x mcp exposes k6 through a built-in Model Context Protocol server, enabling agents to validate scripts, inspect results, and iterate quickly.

Additionally, k6 x docs gives developers CLI access to documentation and API references without leaving the session, while k6 x explore allows browsing the extension registry from the terminal. These features streamline the loop between code generation and validation, a critical need as AI accelerates development.

Beyond AI integrations, k6 2.0 broadens Playwright compatibility within its browser module and introduces a new Assertions API for clearer, more expressive test expectations. Scripts, checks, thresholds, scenarios, and CI/CD workflows remain core to the experience, ensuring existing users feel right at home.

Background

k6, which has amassed over 30,000 stars on GitHub, has long been a go-to tool for proactive performance testing. In 2025, k6 1.0 brought TypeScript support, native extensions, and production-grade stability guarantees. Now, with 2.0, the project responds to a landscape where AI coding assistants are becoming ubiquitous, raising the bar for how quickly and reliably performance tests can be authored and automated.

“The OSS community has embraced k6 because it makes performance testing an integral part of the dev loop,” said James O'Brien, open-source contributor and performance engineer at a major e-commerce platform. “k6 2.0 takes that further by giving agents the same power that humans have—creating tests from natural language prompts and inspecting results in real time.”

What This Means

For engineering teams, k6 2.0 represents a shift toward agentic testing—where AI assistants not only write code but also validate performance automatically. The new k6 x commands lower the barrier to entry for automated testing, enabling developers to go from requirements to a full test suite without leaving their IDE or terminal. This is particularly valuable as organisations adopt AI code generation and need to maintain quality without manual overhead.

The enhanced browser module and Assertions API also improve test reliability for modern web applications. Combined, these features position k6 to support faster, more automated software delivery lifecycles while retaining the simplicity that made it popular. “Teams can now embed performance testing into every stage of development, from a developer’s local machine to production canary runs,” added Liu.

k6 2.0 is available immediately for download from the official website and GitHub repository. A dedicated talk from GrafanaCON 2026 provides further technical deep-dives.