Lunary is a clean LLM analytics tool with a focus on cost. AgentLens covers cost too — and adds quality scoring, hallucination detection, agent waterfalls, GDPR workflows, and verifiable air-gap mode.
| Capability | AgentLens | Lunary |
|---|---|---|
| Cost analytics + per-model breakdown | ✓ | ✓ |
| Self-hosted deployment | ✓ open-core, BSL 1.1 | Apache-2.0 |
| Multi-step agent waterfalls | ✓ trace_agent + spans | Limited |
| Automatic quality + hallucination flags | ✓ heuristic, zero config | Manual eval |
| Air-gap mode (single env var) | ✓ tcpdump-verified | Manual |
| EU-hosted managed cloud (default) | ✓ Frankfurt | Multi-region |
| Audit log + retention + right-to-erasure | ✓ Scale plan | Limited |
| Production support SLA | ✓ Scale+ | Enterprise add-on |
| Framework-agnostic SDK | ✓ | ✓ |
Based on public Lunary pricing as of May 2026. Verify at https://lunary.ai/pricing.
Both tools share a similar mental model. The SDK surface is different but migration is mechanical.
Both cover cost + basic tracing. AgentLens adds quality scoring, hallucination detection, agent waterfalls, and GDPR workflows. Lunary is leaner; AgentLens is more comprehensive.
Yes — same span/trace mental model. ~30 min mechanical migration.
Compliance posture, agent debugging depth, verifiable air-gap, EU-first managed cloud.
15-minute migration call. We'll map your Lunary setup to AgentLens and get you running.