Skip to content
SPECIFICTION
Get listed

Opik

comet.com

From $19/mo Free plan Open source WebWindowsmacOSLinuxAPISelf-hosted
Visit comet.com
Spec score 8.9 #2 of 29 in MLOps Platforms

What it is

Opik helps teams building applications or agents that call an LLM log actions and evaluate behavior across development, testing, and production. It captures and visualizes traces, supports annotation, and provides audit logs. Its evaluation tools include more than 30 metrics for outcomes such as answer relevance, context precision, task completion, and hallucination. In production, Opik can evaluate traces in real time, alert when criteria fail, and track token use and model costs. Test Suites offer pass/fail assertions, an Agent Playground supports end-to-end tests, and a Prompt Optimizer provides six algorithms. Core observability and evaluation features are available as open-source code for local use. Opik includes a REST API and Python client, with Python and TypeScript SDKs listed, and integrations with more than 40 frameworks, model providers, and gateways. Free Cloud allows up to 10 team members, 25,000 spans per month, and 60-day retention. Pro Cloud costs $19 a month and lists up to 50 team members and 100,000 spans monthly.

Spec sheet

What it does
Opik records agent interactions, retrieval, tool calls and model responses, then evaluates and monitors them across development, testing and production.Read 21 Oct 2026 · comet.com
Purpose
Opik logs agent actions and provides automated evaluation workflows to find and fix errors across development, testing, and production.Read 30 Sep 2026 · comet.com
Audience
The pricing page describes Opik as a platform for teams building applications or agents that make calls to an LLM, and says its free plans suit development and testing.Read 30 Sep 2026 · comet.com
Integrations
The integration catalog covers TypeScript, Python, Java, .NET and Ruby ecosystems plus no-code tools and OpenTelemetry.Read 21 Oct 2026 · comet.com
Support
The pricing page lists community and self-service support across plans, email support for Pro and Enterprise, and dedicated support and SLAs for Enterprise.Read 30 Sep 2026 · comet.com
License
The repository and full platform are released under the Apache-2.0 license.Read 21 Oct 2026 · github.com
Headquarters
New York City, United StatesRead 23 Sep 2026 · comet.com
Founded
2017Read 23 Sep 2026 · comet.com
Agent optimization
Opik includes prompt and tool optimization with multiple optimization algorithms and an optimization dashboard.Read 21 Oct 2026 · comet.com
Annotation
Reviewers can annotate traces or complete conversations using custom feedback schemas and annotation queues.Read 21 Oct 2026 · comet.com
API and SDKs
Opik offers a REST API, Python and TypeScript SDKs, and OpenTelemetry ingestion for languages including Java, Ruby and .NET.Read 21 Oct 2026 · comet.com
Company profile
Comet is a privately held software-development company headquartered in New York with 51–200 employees and a 2017 founding year.Read 21 Oct 2026 · linkedin.com
Company usage claim
Comet says more than 150,000 builders use Opik.Read 21 Oct 2026 · linkedin.com
Documentation
Documentation includes feature guides, integration pages, self-hosting instructions, SDK references, REST API references, cookbooks and a changelog.Read 21 Oct 2026 · comet.com
Enterprise security
The pricing page lists enterprise SSO and SOC 2, ISO 27001, ISO 9001, HIPAA, and GDPR compliance for the Enterprise plan.Read 30 Sep 2026 · comet.com
Evaluation
Opik supports datasets, experiments, assertions and LLM-as-a-judge metrics for issues such as hallucination, relevance and moderation.Read 21 Oct 2026 · github.com
Framework integrations
Named integrations include LangChain, LangGraph, LlamaIndex, CrewAI, AutoGen, Google ADK, OpenAI Agents, Semantic Kernel and many others.Read 21 Oct 2026 · comet.com
Free plan limits
The Free Cloud plan lists up to 10 team members, 25,000 spans per month, and 60-day data retention.Read 30 Sep 2026 · comet.com
Funding
TechCrunch reported a $50 million Series B in November 2021, bringing Comet's reported total raised to almost $70 million.Read 21 Oct 2026 · techcrunch.com
Guardrails
Self-hosted deployments can use PII, topic and custom guardrails to detect or block unsafe inputs and outputs.Read 21 Oct 2026 · comet.com
Launch
Comet announced Opik as an open-source LLM evaluation platform in September 2024.Read 21 Oct 2026 · comet.com
Limit: interface
The Slashdot reviewer said the user interface could be made more attractive.Read 21 Oct 2026 · slashdot.org
Limit: learning curve
The Gartner reviewer said it took time to fully understand the product.Read 21 Oct 2026 · gartner.com
MCP extensibility
Its MCP server lets Claude Code, Cursor, VS Code Copilot, Codex and opencode query traces, score outputs and run evaluations from chat.Read 21 Oct 2026 · github.com
Open source
The core observability and evaluation feature set is available as open-source code that users can download and run locally.Read 30 Sep 2026 · comet.com
Praise: debugging
The Gartner reviewer praised feedback scores, cost visibility and debugging features for improving AI output quality.Read 21 Oct 2026 · gartner.com
Praise: feature breadth
The Slashdot reviewer praised support for many LLM providers, scalability for high trace loads and a fast-moving roadmap.Read 21 Oct 2026 · slashdot.org
Pro plan limit
The Pro Cloud plan lists up to 50 team members and 100,000 spans per month, with customizable span limits and data retention periods.Read 30 Sep 2026 · comet.com
Production monitoring
Online evaluation rules can score incoming traces while dashboards track feedback, latency, cost and errors.Read 21 Oct 2026 · comet.com
Prompt and agent testing
Opik offers Test Suites with pass/fail assertions, an Agent Playground for end-to-end tests, and a Prompt Optimizer with six algorithms.Read 30 Sep 2026 · comet.com
Prompt management
Its prompt tools provide a library, version history, playground, composable prompt building and dataset-based comparisons.Read 21 Oct 2026 · comet.com
Provider integrations
Named model-provider integrations include Anthropic, Amazon Bedrock, Gemini, Cohere, DeepSeek, Fireworks, Groq, Mistral, Ollama, OpenAI and xAI.Read 21 Oct 2026 · comet.com
Rating on G2
No rating; 0 reviews listed for OpikRead 21 Oct 2026 · g2.com
Rating on Gartner Peer Insights
4.0 out of 5 from 1 reviewRead 21 Oct 2026 · gartner.com
Rating on Slashdot
5.0 out of 5 from 1 reviewRead 21 Oct 2026 · slashdot.org
Release activity
The changelog records an Ollama integration released on October 12, 2026, including streaming and token-usage tracing.Read 21 Oct 2026 · comet.com
Security and compliance
Enterprise pricing lists SSO, service accounts, view-only users, flexible deployments and SOC 2, ISO 27001, ISO 9001, HIPAA and GDPR compliance.Read 21 Oct 2026 · comet.com
Self-hosting
Self-hosting exposes the full feature set but omits user-management features; local Docker installation is for getting started, while Kubernetes is the production deployment option.Read 21 Oct 2026 · comet.com
System requirements
The documented local setup assumes Git and Docker; Linux, macOS and Windows installation scripts are provided.Read 21 Oct 2026 · comet.com
Tracing
The platform captures and visualizes multi-step agent execution, including LLM calls, tools and retrieval steps.Read 21 Oct 2026 · comet.com
Tracing and debugging
Opik captures and visualizes agent actions, supports trace annotation, and provides audit logs.Read 30 Sep 2026 · comet.com

Plans and prices

Free Cloud Free Free plan Up to 10 team members · 25k spans/month · 60-day retention Read 21 Oct 2026 · comet.com
Open Source Free Free Full AI observability and agent testing · self-hosted Read 21 Oct 2026 · comet.com
Pro Cloud $19 a month Per month Up to 50 team members · 100k spans/month · 60-day retention · customizable span limits and retention Read 21 Oct 2026 · comet.com
Enterprise Priced on request Custom Unlimited team members · custom usage plans · flexible deployments · SSO · dedicated support and SLAs · SOC 2, ISO 27001, ISO 9001, HIPAA and GDPR Read 21 Oct 2026 · comet.com

Every Opik plan and price →

Strong points and limits

Strong points

  • Includes more than 30 evaluation metrics.
  • Can evaluate production traces in real time.
  • Tracks token usage and model costs.
  • Open-source code can be run locally.
  • Free Cloud includes 25,000 spans monthly.

Limits

  • Free Cloud is limited to 10 team members.
  • Free Cloud retains data for 60 days.
  • Pro Cloud costs $19 a month.
  • Enterprise pricing is custom and not listed.

Who it suits

Opik suits teams developing applications or agents that make calls to an LLM and want tracing and evaluation across the development cycle. Its free plans are described as suitable for development and testing.

Opik in depth

Read the full write-up

Opik is a platform for tracing, evaluating, testing, and monitoring applications and agents that call LLMs. It best suits teams that want to follow agent behavior from development into production, including teams that prefer to run open-source software themselves. Its broad workflow is a strong draw; hosted users should weigh monthly span limits against expected volume.

Overview

Opik connects observability with evaluation and testing rather than treating production monitoring as a separate task. Teams can inspect agent actions, evaluate outcomes, test prompts and agents, then monitor production traces for failures and costs. The breadth is useful for teams building LLM applications, but it also means cloud plan limits deserve attention as usage grows.

Key features

Tracing captures and visualizes agent actions, with annotation and audit logs for investigation and review. Evaluation includes more than 30 metrics spanning answer relevance, context precision, task completion, hallucination, and other outcomes. That gives teams a broad set of measures to work with, though they still need to choose criteria that reflect their application.

Production monitoring can evaluate traces in real time, alert when criteria fail, and track token usage and model costs. For development, Test Suites provide pass/fail assertions, the Agent Playground supports end-to-end tests, and the Prompt Optimizer includes six algorithms. This links pre-release checks to ongoing monitoring, a useful fit for teams that want evaluation to continue after deployment.

The core observability and evaluation feature set is open source and can be run locally. Opik includes a REST API and Python client; its pricing page also lists Python and TypeScript SDKs. More than 40 integrations cover AI frameworks, model providers, and gateways, with LangChain, OpenAI, Google ADK, and LangGraph among the named examples.

Pricing

Opik uses a freemium model, with hosted plans and a free self-hosted option. Free Cloud costs nothing and is billed $0. It allows up to 10 team members, 25,000 spans per month, and 60-day data retention. Those caps make it a practical starting point for development and testing, but teams with heavier trace volume or longer retention needs will need another plan.

Open Source costs nothing and is billed $0. It provides the full AI observability and agent testing feature set, uses the same codebase as the hosted versions, and is self-hosted. It avoids the stated cloud span quota, but requires teams to run the software themselves.

Pro Cloud costs $19 a month, billed $19 per month. It raises the allowance to up to 50 team members and 100,000 spans per month, with 60-day retention and customizable span limits and retention periods. Pro is the natural step up for teams that have outgrown Free Cloud; the listed base allowance still makes it important to match plan limits to workload.

Enterprise has custom pricing. It includes unlimited team members, custom usage plans, flexible deployments, dedicated support, and SLAs. Enterprise also lists SSO and compliance with SOC 2, ISO 27001, ISO 9001, HIPAA, and GDPR, making it the tier aimed at organizations with those requirements. Community and self-service support are listed across plans, while Pro and Enterprise add email support.

Platforms

Opik is available through an API and on the web, and supports Linux and self-hosted deployment. The choice of hosted cloud or self-hosting accommodates teams with different deployment preferences.

Who it's for

Opik is aimed at teams building applications or agents that make calls to an LLM. Its free plans suit development and testing, while the hosted plans offer a managed option with defined team, span, and retention limits. The local open-source version is a stronger fit for teams prepared to self-host and wanting the full observability and agent testing feature set.

Pros and cons

Pros: Tracing, evaluation, testing, and production monitoring sit in one workflow, including real-time criteria alerts and token-cost tracking.

Pros: The open-source edition provides the full stated feature set for self-hosting at free.

Pros: More than 30 metrics and over 40 integrations give teams a broad starting point for evaluation and connecting existing AI tools.

Cons: Free Cloud is capped at 25,000 spans per month and 60-day retention, which may constrain teams with higher-volume or longer-term needs.

Cons: Pro Cloud's $19-per-month plan still has a 100,000-span monthly allowance; higher usage or customized retention may require changes to the plan.

Cons: Self-hosting keeps the open-source option free, but makes operating the software the team's responsibility.

Alternatives

W&B Weave is worth considering for teams that want a hosted free tier with AI application evaluations, tracing, and scorers; its free plan specifies 1 GB/mo of data ingestion and 5 GB/mo of storage.

OpenLIT is an alternative for readers seeking a free option whose OSS plan specifies unlimited self-hosted usage, users, projects, and environments.

Helicone is another freemium, open-source alternative.

Maxim AI may suit a small team looking for a free developer plan, which allows up to 3 seats, 10k logs per month, and 3-day data retention.

Respan (formerly Keywords AI) offers a free plan with 100k logs and 1k scores, among other allowances, for teams comparing usage quotas.

Agenta is an alternative with a free Hobby plan capped at 2 team members, 5,000 agent runs per month, 20 evaluations per month, and one-week trace retention.

LangWatch is an open-source alternative under Apache 2.0; its Developer plan lists 50k events per month and 14-day data access.

Portkey is another freemium option, with a Developer plan that lists 10k recorded logs per month and three days of log retention.

For broader comparisons, see AI LLM Observability Tools, LLM Observability Tools, Model Monitoring Software, AI Agent Observability Tools, AI Agent Evaluation Tools, and Machine Learning Model Monitoring Software.

Verdict

Opik is a strong choice for teams that need an end-to-end workflow for LLM agent tracing, evaluation, testing, and production monitoring, especially if self-hosting matters. Its open-source edition gives teams the full feature set without a plan fee, while hosted users get a clear route from a free tier to Pro and custom Enterprise usage. Look elsewhere if your cloud workload does not fit the span ceilings or if self-hosting is not appealing and the hosted allowances are too restrictive.

Verdict

Opik covers tracing, evaluation, testing, and production monitoring, with both hosted plans and an open-source local option. Check span and retention limits against your team's workload before selecting a plan.

Where Specifiction places it

Alternatives to OpikAll alternatives →