Keploy

Keploy is a traffic-driven testing platform for teams that want regression tests and dependency sandboxes based on real application requests. Its strongest fit is API and service teams working across local development, CI, and Kubernetes. The open-source local recorder and a usable free Playground make it approachable; Pro raises the quotas, while Kubernetes recording and production capture require custom-priced Enterprise.

Overview

Keploy captures incoming requests and outgoing dependency calls through eBPF and network-level interception, without changes to application code. It replays that traffic as regression tests in CI, using mocks and isolation. This approach is useful when teams want to preserve real request and dependency behavior in a test suite, rather than construct every case manually. SOAP is unsupported, so it is a poor fit for teams whose API coverage depends on that protocol.

Key features

Capture, generate, replay

Recorded traffic can be reused to catch regressions, while the Playground plan can generate API tests from OpenAPI or Postman inputs and use AI to expand edge cases. That gives teams a way to build an initial suite from existing API definitions, but the free plan's five AI credits per month and 30 suite generations limit repeated use. Pro raises those caps to 20 credits and 100 generations monthly.

Protocols and integrations

Keploy supports HTTP(S), HTTP/2, gRPC, Protobuf, GraphQL, and Kafka, plus databases including MySQL, MongoDB, Redis, DynamoDB, Postgres, and RDS. GitHub, GitLab, Jenkins, and Argo integrations connect it to common CI/CD workflows; AWS Marketplace is also available. The breadth makes it relevant to teams testing services with varied dependencies, though the published protocol coverage does not include SOAP.

Browser capture and security

The Chrome extension captures XHR and fetch traffic with headers, cookies, and authentication context, then exports calls as cURL, JSON, or Keploy YAML. That can help turn browser-driven API activity into reusable cases. Keploy describes access controls, encryption at rest and in transit, regular audits, and anonymization where possible for AI processing. Enterprise adds SOC 2, GDPR, and HIPAA compliance features, plus cloud, self-hosted, and air-gapped deployment options.

Pricing

The Open Source plan is free and self-hosted, with local recording on macOS, Windows, Linux, and Docker, plus CI replay. It suits developers who can run the tooling themselves and need the core record-and-replay workflow. The core platform is open source under Apache 2.0.

Playground is free forever and includes 30 test suite generations, 100 test runs, 5,000 integration and sandbox runs, and five AI credits per month. It is a solid starting point for individual evaluation or modest usage, but its quotas may constrain teams running larger suites or generating tests regularly.

Pro costs $19/user/mo plus additional usage. It includes 100 suite generations, 400 test runs, 20,000 integration and sandbox runs, 20 AI credits, and email and chat support. It suits teams that have outgrown Playground's caps and want direct support; usage beyond the included quotas adds cost.

Enterprise has custom pricing and includes production traffic capture, Kubernetes recording, SSO and RBAC, private or air-gapped deployments, a dedicated support engineer, and onboarding. Pro CI replay uses a Keploy API key; Enterprise can replay in a Kubernetes cluster or in CI using a Keploy API key. Enterprise is the relevant tier for organizations that need its deployment, access-control, or production capture options.

Platforms

Keploy lists API, browser extension, Linux, macOS, Windows, web, and self-hosted availability. The Open Source plan specifically supports local recording on macOS, Windows, Linux, and Docker.

Who it's for

Keploy fits developers and service teams that want to turn existing traffic into repeatable regression tests, mock dependencies, or generate API suites from OpenAPI and Postman definitions. Its free local option works for self-hosting developers, while Playground and Pro suit teams whose run volume fits their quotas. Teams that need SOAP coverage or production-focused capabilities without moving to custom-priced Enterprise should consider another tool.

Pros and cons

• Pros: Captures requests and dependency calls without application code changes, making existing traffic useful as test input.
• Pros: Free local record-and-replay and a free Playground provide two ways to start, including API test generation from OpenAPI or Postman.
• Pros: Broad listed protocol and database support, plus GitHub, GitLab, Jenkins, and Argo integrations, suits varied service stacks.
• Cons: SOAP is not supported, ruling it out for teams that need SOAP testing.
• Cons: Playground and Pro enforce monthly generation, run, sandbox, and AI-credit quotas; additional Pro usage costs extra.
• Cons: Production capture, Kubernetes recording, and private or air-gapped deployment are Enterprise features with custom pricing.

Alternatives

For a fully open-source, community-driven option, consider Microcks. If you need endpoint mock rules and CRUD objects, Beeceptor offers an Individual plan at $10 a month with 15,000 requests per month per endpoint and up to 50 mock rules. Teams focused on emulating AWS services can choose LocalStack, whose free Hobby plan includes 30+ emulated services and one personal sandbox for non-commercial use.

MockServer is a free, self-hosted option with no SaaS account requirement. For open-source contract testing, consider Pact. Specmatic is worth considering if you need WSDL and SOAP alongside OpenAPI, contract testing, or service virtualization. Testcontainers Desktop offers individual developers 300 minutes of cloud runtime per month on its free plan. Karate covers REST, GraphQL, and SOAP API testing, as well as browser and desktop UI automation and Gatling performance testing.

Browse more Integration Testing Tools, AI Software Testing Tools, and API Testing Tools.

Verdict

Choose Keploy if your team wants to turn captured application and dependency traffic into regression tests, with local open-source tooling and free hosted quotas to get started. Its protocol range and traffic-based workflow are compelling for API and service testing. Look elsewhere if SOAP is essential, or if you need production capture and Kubernetes recording but cannot justify Enterprise custom pricing.