HoneyHive

What it does

HoneyHive is an agent observability and evaluation platform. It ingests traces of LLM calls, tool calls, and handoffs sent through OpenTelemetry or its Python and TypeScript SDKs, and turns them into step-by-step trace replays, live quality scores, alerts, experiment comparisons, and datasets built from production failures.

Teams use it to inspect what an agent did in a run, score live traffic with LLM judges or code evaluators, route flagged traces to human annotation queues, and run automated evals in CI on every release. Coding agents can use a CLI, an MCP server, and purpose-built skills to investigate alerts.

It suits AI engineers and platform teams instrumenting custom or third-party agents, plus risk and governance teams needing audit evidence. The Developer plan is free with 10,000 events per month, up to five users, and 30-day retention. Enterprise pricing is custom and adds SAML and custom SSO, higher limits, and optional self-hosting.

What it can do

• Trace agent runs step by step, including LLM calls, tool calls, and handoffs
• Run automated evaluations using LLM judges or code-based evaluators on live traffic
• Set alerts on score thresholds with email or webhook notifications
• Compare agent versions on the same dataset to detect regressions
• Route flagged traces to human annotation queues
• Export traces, evaluations, and metrics to data warehouses in open formats

Works with

• OpenTelemetry
• LangChain
• LangGraph
• AWS Strands
• Google ADK
• OpenAI Agents SDK
• GitHub Actions
• Snowflake
• Databricks
• BigQuery

Before you choose

• Free Developer plan is capped at 10,000 events per month and 30-day retention, with up to 5 users and 1 workspace.
• Self-hosting, SAML/custom SSO, and custom data retention are Enterprise-only.
• Ingestion keys send traces and events only; running experiments or reading data back requires a separate project API key.

Listed under

• Coding
• DevOps and observability
• Agent tracing
• AI observability
• Developers
• Enterprise AI
• LLM evaluation
• LLM monitoring

Get started

Create a project and ingestion API key, then install the SDK and trace a first call.

1. Create a project in HoneyHive and generate an ingestion API key from project settings.
2. Install the honeyhive SDK and an instrumentor for your model provider (Python 3.11+ required in the quickstart).
3. Initialize HoneyHiveTracer and the instrumentor, run an instrumented call, then view the trace in the Traces page.

The Developer plan is free with 10,000 events per month, a recurring free entitlement rather than a time-limited trial. Paid rates above that were not confirmed on the page; Enterprise is custom pricing requiring a conversation with the team.

Source: honeyhive.ai/pricing, checked 29 Oct 2026. Prices can change; check before you buy.

Sources for this overview

Product pages, documentation, and pricing checked 29 Oct 2026. Each source supports part of the overview; availability and prices can change.

• HoneyHive - Agent Observability & Evaluation Platform ↗www.honeyhive.ai
• How to Trace Your First Agent - HoneyHive AI Docs ↗docs.honeyhive.ai
• HoneyHive Pricing ↗www.honeyhive.ai

Company website

honeyhive.ai: product, pricing and policy pages

YC directory mirror

Batch and company name

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

manual_review

Last checked 29 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.