Keploy

Keploy is an open-source test generation platform that uses production traffic and eBPF to create isolated sandboxes for API and unit testing.

Keploy records API traffic, dependency calls, and data states to build a digital twin of production behavior. It uses eBPF and network-level interception to capture this data without requiring application code changes. Developers can replay these recorded interactions in local or CI/CD environments to identify regressions in responses, schemas, and performance. The platform also generates additional edge cases and scenarios using API schemas and application context. It serves engineering teams by automating the creation of integration tests and mocks for databases, queues, and external APIs. Keploy is compatible with Kubernetes, local environments, and existing CI/CD pipelines.

Keploy at a glance

Best for: Software engineer
Runs on: Kubernetes, CLI, open source, self-hosted
Pricing: Free Playground tier; Pro $24/user/month; Enterprise plan available with custom pricing.
Released: January 2022
Category: AI coding tool
Website: keploy.io

Key features

Automatic test generation from live traffic
Dependency mocking for databases, queues, and APIs
eBPF-based network traffic interception
Human-editable YAML test configurations
AI-powered test generation and self-healing
Support for local or CI/CD test execution
Contract and load testing capabilities

Use it for

Generating API regression tests from real production traffic
Creating isolated sandbox environments for CI/CD
Mocking external dependencies during development
Detecting performance and schema regressions before deployment

Alternatives to Keploy

Tusk — Generates tests from production traffic to catch regressions in PRs.
Coroot — Uses eBPF for zero-instrumentation monitoring with minimal overhead.
Genkit — Provides unified APIs and local developer tools for AI application workflows.
KushoAI — Autonomous agents that self-heal tests within CI/CD pipelines.
CopilotKit — Connects your application to AI agents with generative UI and human approvals.
Open_SwarmForce — It combines automated code review with deterministic test verification in sandboxes.
VulnClaw — It uses LLM agents to autonomously execute end-to-end security testing loops.
TestSprite — Runs end-to-end tests on live apps to catch regressions without manual scripting.
ARTEX — Runs autonomous multi-agent penetration testing with a dual-graph architecture.
PentAGI — It autonomously manages penetration testing workflows within a secure Docker sandbox.
OpenCode — Works with 75+ model providers, local models, and existing subscriptions
OpenSandbox — Provides isolated environments with multi-language SDKs and lifecycle management.

All Keploy alternatives