Gretel.ai vs Hoverfly

Free tier available

Free and open source

Consumption-based on generated records; enterprise agreements available.

Free and open source. You pay in engineering time and CI minutes.

Free tier with a monthly credit allowance.

Everything, self-hosted

Emerging

Gretel (NVIDIA) · since 2019

Emerging

SpectoLabs · since 2016

• Formal differential-privacy guarantees rather than best-effort masking.
• Synthetic data preserves statistical relationships useful for analytics and model training.
• Privacy and quality scores make the trade-off explicit.

• Capture mode produces a realistic simulation from live traffic with no authoring.
• Latency injection makes it useful for performance and resilience testing too.
• Small footprint — a single Go binary.

• Model training adds time and cost compared to rule-based masking.
• Overkill when simple anonymisation would satisfy the requirement.
• NVIDIA acquisition changes the strategic outlook.

• Smaller community and slower development than WireMock.
• Proxy configuration in complex networks takes effort.
• Captured simulations need periodic refresh.

Regulated organisations that need shareable data with defensible privacy properties.

Teams needing lightweight service virtualisation against real recorded traffic.