CM

Comet ML

Infrastructure

MLOps platform for experiment tracking, model production monitoring and LLM evaluation.

Total funding$70M

Valuation—

Employees80

Founded2017

HeadquartersNew York, USA

StatusPrivate

Products

Selected offerings — the AI products recorded here, not the vendor's full catalogue.

• Opik Platform Sold on its own

  GenAI observability and agent-testing platform from Comet, for tracing, evaluating and debugging LLM applications.

• Comet Experiment Tracking Developer tool

  Comet's experiment-tracking product for logging, comparing, visualizing and reproducing machine-learning training runs.

• Comet Artifacts Developer tool

  Comet's dataset- and model-versioning product for tracking artifacts and lineage from training through production.

• Comet Model Production Monitoring Platform

  Comet's production-monitoring product for tracking deployed-model performance and data drift across the ML lifecycle.

Inside Opik

• Agent Optimizer Developer tool Included with the parent

  Opik feature (also available as an open-source SDK) that automatically iterates and tunes system prompts across seven optimization algorithms before freezing them for production.

• Cost Intelligence Application

  Opik feature that tracks and attributes AI/coding-agent token spend (Claude Code, Codex) across engineering teams, surfacing recoverable-spend opportunities.

Similar products 94

No sourced competitor yet: no filing, vendor document or measured-market report read so far names a rival for this company, so none is asserted. The companies below are the closest by what their products do.

How this list is ranked ↗

W&Weights & BiasesAcquiredOpik and W&B Weave both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; Weights & Biases is acquired, with no independent scale.2 of 4 capabilitiesShared product typeCompareMModalPrivate · $466M raisedComet Artifacts and Modal Notebooks do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private).1 of 4 capabilitiesShared product typeCompareCCrusoePrivate · $10BComet Artifacts and Serverless Fine-Tuning do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private).2 of 4 capabilitiesShared product typeCompareDDatabricksPrivate · $190BComet Experiment Tracking and Databricks Model Training do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private).3 of 4 capabilitiesShared product typeCompareDDeepchecksAbsorbedComet Model Production Monitoring and Deepchecks LLM Evaluation both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; Deepchecks is absorbed, with no independent scale.1 of 4 capabilitiesShared product typeCompareTATogether AIPrivate · $8.3BComet Artifacts and Together Custom Training do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private).1 of 4 capabilitiesShared product typeCompareCCoreWeavePublic · $48.5BCost Intelligence and CoreWeave ARIA both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; similar scale (public).3 of 4 capabilitiesShared product typeCompareNGNebius GroupPublic · $61.6BComet Artifacts and Managed Service for MLflow do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (public).2 of 4 capabilitiesShared product typeCompareBBasetenPrivate · $13BComet Experiment Tracking and Baseten Training do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private).1 of 4 capabilitiesShared product typeCompareFAFireworks AIPrivate · $17.5BComet Artifacts and Fireworks Training do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private); ships API service and model API rather than the same layer.1 of 4 capabilitiesDifferent layerCompareSSplunkAcquiredComet Model Production Monitoring and Splunk Observability Cloud both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; Splunk is acquired, with no independent scale.3 of 4 capabilitiesShared product typeCompareFAFiddler AIPrivate · $100M raisedComet Model Production Monitoring and Fiddler AI both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; similar scale (private).1 of 4 capabilitiesShared product typeCompareBBraintrustPrivate · $125M raisedComet Model Production Monitoring and Braintrust both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; similar scale (private).1 of 4 capabilitiesShared product typeCompareGGalileoAcquired ($68M deal)Comet Model Production Monitoring and Galileo both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; Galileo is acquired, with no independent scale.1 of 4 capabilitiesShared product typeCompareUUnslothPrivateComet Artifacts and Unsloth do comparable work on model training; both also serve buyers who need to train or fine-tune a model; Unsloth's scale not recorded.1 of 4 capabilitiesShared product typeCompareAAArthur AIPrivate · $62.7M raisedComet Model Production Monitoring and Arthur both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; same layer and the same scale band (early-stage private).1 of 4 capabilitiesShared product typeSame scale bandCompareTTuringPrivate · $2.2BComet Artifacts and RL environments do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private); ships agent platform and data service rather than the same layer.2 of 4 capabilitiesDifferent layerCompareTTolokaPrivateComet Artifacts and Toloka Train do comparable work on model training; both also serve buyers who need to train or fine-tune a model; Toloka's scale not recorded.3 of 4 capabilitiesShared product typeCompareLLabelboxPrivate · $189M raisedComet Artifacts and Terra do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private).3 of 4 capabilitiesShared product typeCompareEEleutherAINonprofitComet Artifacts and GPT-NeoX do comparable work on model training; both also serve buyers who need to train or fine-tune a model; EleutherAI is a nonprofit, with no independent scale.2 of 4 capabilitiesShared product typeCompareAAArize AIPrivateOpik and Arize Phoenix both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; Arize AI's scale not recorded.2 of 4 capabilitiesShared product typeCompareCCCactus ComputePrivateComet Artifacts and Cactus Platform do comparable work on model training; both also serve buyers who need to train or fine-tune a model; Cactus Compute's scale not recorded.1 of 4 capabilitiesShared product typeCompareNNeo4jPrivate · $2BComet Artifacts and Neo4j Graph Data Science do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private).2 of 4 capabilitiesShared product typeCompareLLangfuseAcquiredOpik and Langfuse both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; Langfuse is acquired, with no independent scale.1 of 4 capabilitiesShared product typeCompareHHeliconeAcquiredOpik and Helicone both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; Helicone is acquired, with no independent scale.1 of 4 capabilitiesShared product typeCompareNCNAVER CloudPrivateComet Artifacts and NCLUE do comparable work on model training; both also serve buyers who need to train or fine-tune a model; NAVER Cloud's scale not recorded.2 of 4 capabilitiesShared product typeCompareAAppenPublicComet Artifacts and AI Data Platform do comparable work on model training; both also serve buyers who need to train or fine-tune a model; Appen's scale not recorded.1 of 4 capabilitiesShared product typeCompareHFHugging FacePrivate · $4.5BComet Experiment Tracking and Hugging Face Jobs do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private).1 of 4 capabilitiesShared product typeCompareHCHuawei CloudPrivateComet Experiment Tracking and Huawei HiLens do comparable work on model training; both also serve buyers who need to train or fine-tune a model; Huawei Cloud's scale not recorded.3 of 4 capabilitiesShared product typeCompareRRunPodPrivateComet Experiment Tracking and Pods do comparable work on model training; both also serve buyers who need to train or fine-tune a model; RunPod's scale not recorded.1 of 4 capabilitiesShared product typeCompareCClickHousePrivate · $1.1B raisedOpik and ClickStack both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; similar scale (private).2 of 4 capabilitiesShared product typeCompareIIBMPublic · $216.6BComet Model Production Monitoring and IBM watsonx.governance both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; similar scale (public).4 of 4 capabilitiesShared product typeCompareAAlphabet (Google)Public · $4.2TComet Experiment Tracking and Vertex AI Workbench do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (public).4 of 4 capabilitiesShared product typeCompareBByteDancePrivateOpik and Coze Loop both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; ByteDance's scale not recorded.3 of 4 capabilitiesShared product typeCompareLLambdaPrivate · $1.5B raisedComet Artifacts and Lambda 1-Click Clusters do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private); ships infrastructure service and model API rather than the same layer.1 of 4 capabilitiesDifferent layerCompareOOpenRouterPrivate · $1.3BAgent Optimizer and Ori do comparable work on agentic coding; similar scale (private).1 of 4 capabilitiesShared product typeCompareHHuaweiPrivateComet Artifacts and MindSpore do comparable work on model training; both also serve buyers who need to train or fine-tune a model; Huawei's scale not recorded.2 of 4 capabilitiesShared product typeCompareCSCerebras SystemsPublic · $44.4BComet Experiment Tracking and Cerebras Training Cloud do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (public).1 of 4 capabilitiesShared product typeCompareEElasticPublic · $10.2BOverlaps on data analysis and evaluation and observability; similar scale (public).2 of 4 capabilitiesShared product typeCompareLLangChainPrivate · $160M raisedComet Model Production Monitoring and LangSmith both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; similar scale (private).1 of 4 capabilitiesShared product typeCompareMMicrosoftPublic · $3.7TOpik and Observability in Foundry Control Plane both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; similar scale (public).4 of 4 capabilitiesShared product typeCompareSSnowflakePublic · $116BAgent Optimizer and Snowflake CoCo do comparable work on agentic coding; similar scale (public).2 of 4 capabilitiesShared product typeCompareALAI21 LabsPrivate · $1.4BComet Artifacts and AI21 Post-Training do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private); ships API service, agent platform and 2 more rather than the same layer.1 of 4 capabilitiesDifferent layerCompareNNVIDIAPublic · $5.1TComet Experiment Tracking and NVIDIA AI Workbench do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (public).3 of 4 capabilitiesShared product typeCompareTMThinking Machines LabPrivate · $12BComet Artifacts and Tinker do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private); ships API service rather than the same layer.1 of 4 capabilitiesDifferent layerCompareMAMistral AIPrivate · $13.8BComet Artifacts and Mistral Forge do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private).2 of 4 capabilitiesShared product typeCompareOOraclePublic · $454.4BComet Experiment Tracking and OCI Data Science do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (public).2 of 4 capabilitiesShared product typeCompareDDataikuPrivate · $3.7BComet Model Production Monitoring and Dataiku Agent Hub both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; similar scale (private).3 of 4 capabilitiesShared product typeCompareHAHume AIPrivateOpik and Kairos Platform both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; Hume AI's scale not recorded.2 of 4 capabilitiesShared product typeCompareAMApplied MaterialsPublic · $380.7BComet Model Production Monitoring and SmartFactory AI IQworks both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; similar scale (public).3 of 4 capabilitiesShared product typeCompareSSiemensPublic · $244.7BComet Model Production Monitoring and Industrial AI Suite both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; similar scale (public).3 of 4 capabilitiesShared product typeCompareKKuaishouPublic · $18.6BAgent Optimizer and CodeFlicker do comparable work on agentic coding; similar scale (public).2 of 4 capabilitiesShared product typeCompareBBaiduPublic · $31.6BAgent Optimizer and Baidu Comate do comparable work on agentic coding; similar scale (public).3 of 4 capabilitiesShared product typeCompareSSAPPublic · $256.3BComet Model Production Monitoring and SAP AI Agent Hub both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; similar scale (public).3 of 4 capabilitiesShared product typeCompare44ParadigmPublic · $2BComet Artifacts and 4Paradigm Sage HyperCycle do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (public).1 of 4 capabilitiesShared product typeCompareIMInsilico MedicinePublic · $4.3BComet Artifacts and MMAI Gym do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (public).3 of 4 capabilitiesShared product typeCompareAAdobePublic · $115.9BComet Artifacts and Adobe Firefly Custom Models do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (public).2 of 4 capabilitiesShared product typeCompareAAlibabaPublic · $273.4BComet Experiment Tracking and Alibaba Cloud Platform for AI (PAI) do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (public).2 of 4 capabilitiesShared product typeCompareIIntelPublic · $463.7BComet Experiment Tracking and Intel Geti do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (public).1 of 4 capabilitiesShared product typeCompareSServiceNowPublic · $130.1BComet Model Production Monitoring and AI Control Tower both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; similar scale (public).3 of 4 capabilitiesShared product typeCompareSASnorkel AIPrivate · $1.3BComet Artifacts and Snorkel Flow do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private).1 of 4 capabilitiesShared product typeCompareLALiquid AIPrivateComet Artifacts and LEAP do comparable work on model training; both also serve buyers who need to train or fine-tune a model; Liquid AI's scale not recorded.1 of 4 capabilitiesShared product typeCompareOOpenAIPrivate · $852BComet Model Production Monitoring and OpenAI Frontier both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; similar scale (private).3 of 4 capabilitiesShared product typeCompareYYandexPublicAgent Optimizer and SourceCraft Code Assistant do comparable work on agentic coding; Yandex's scale not recorded.3 of 4 capabilitiesShared product typeCompareAAmazonPublic · $2.8TAgent Optimizer and Amazon Q Developer do comparable work on agentic coding; similar scale (public).4 of 4 capabilitiesShared product typeCompareAAMDPublic · $785.1BComet Experiment Tracking and AMD Developer Cloud do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (public).1 of 4 capabilitiesShared product typeCompareCCognitionPrivate · $26BAgent Optimizer and Fusion do comparable work on agentic coding; similar scale (private).1 of 4 capabilitiesShared product typeCompareIImbuePrivate · $232M raisedAgent Optimizer and Blueprint do comparable work on agentic coding; similar scale (private).1 of 4 capabilitiesShared product typeCompareTTencentPublic · $510.1BAgent Optimizer and CodeBuddy do comparable work on agentic coding; similar scale (public).2 of 4 capabilitiesShared product typeCompareRReflectionPrivate · $27.5BAgent Optimizer and Asimov do comparable work on agentic coding; similar scale (private).1 of 4 capabilitiesShared product typeCompareCCognigyAcquiredComet Model Production Monitoring and Cognigy AI Ops Center both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; Cognigy is acquired, with no independent scale.2 of 4 capabilitiesShared product typeCompareUUiPathPublic · $7.2BComet Artifacts and UiPath AI Center do comparable work on model training; similar scale (public).3 of 4 capabilitiesShared product typeCompareCCursorAcquired ($60B deal)Agent Optimizer and Bugbot do comparable work on agentic coding; Cursor is acquired, with no independent scale.1 of 4 capabilitiesShared product typeCompareBFBlack Forest LabsPrivate · $3.3BComet Experiment Tracking and Open Weights Licensing do comparable work on model training; both also serve buyers who need to train or fine-tune a model; similar scale (private).1 of 4 capabilitiesShared product typeCompareDDXPrivateCost Intelligence and AI Code Insights both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; DX's scale not recorded.2 of 4 capabilitiesShared product typeCompareHHumainPrivateComet Artifacts and HUMAIN COMPUTE do comparable work on model training; both also serve buyers who need to train or fine-tune a model; Humain's scale not recorded; ships API service, agent platform and 1 more rather than the same layer.2 of 4 capabilitiesDifferent layerCompareFFreshworksPublic · $3.4BAgent Optimizer and Freddy AI Copilot for Developers do comparable work on agentic coding; similar scale (public).2 of 4 capabilitiesShared product typeCompareGGensparkPrivate · $645M raisedAgent Optimizer and GenTerminal do comparable work on agentic coding; similar scale (private).2 of 4 capabilitiesShared product typeCompareSSonarPrivate · $4.7BAgent Optimizer and SonarQube do comparable work on agentic coding; similar scale (private).1 of 4 capabilitiesShared product typeCompareFFinAcquired ($3.6B deal)Cost Intelligence and Fin Analyze both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; Fin is acquired, with no independent scale.2 of 4 capabilitiesShared product typeCompareDDeepSeekPrivate · $50BAgent Optimizer and DeepSeek Harness do comparable work on agentic coding; similar scale (private).1 of 4 capabilitiesShared product typeCompareRRunwayPrivate · $5.3BComet Artifacts and Runway Video Model Licensing do comparable work on model training; similar scale (private).1 of 4 capabilitiesShared product typeCompareYYellow.aiPrivateCost Intelligence and Analyze both serve buyers who need to watch an AI agent for suspicious or runaway behaviour; Yellow.ai's scale not recorded.2 of 4 capabilitiesShared product typeCompareVVercelPrivate · $9.3BAgent Optimizer and Vercel Agent do comparable work on agentic coding; similar scale (private).2 of 4 capabilitiesShared product typeCompareJJetBrainsPrivateAgent Optimizer and JetBrains AI do comparable work on agentic coding; JetBrains's scale not recorded.1 of 4 capabilitiesShared product typeCompareSShopifyPublic · $196.7BAgent Optimizer and Shopify AI Toolkit do comparable work on agentic coding; similar scale (public).2 of 4 capabilitiesShared product typeCompareZZ.ai (Zhipu AI)Public · $58.6BAgent Optimizer and GLM Coding Plan do comparable work on agentic coding; similar scale (public).2 of 4 capabilitiesShared product typeCompareGGreptilePrivateAgent Optimizer and TREX do comparable work on agentic coding; Greptile's scale not recorded.1 of 4 capabilitiesShared product typeCompareSSpace42Public · $2.6BComet Artifacts and GIQ do comparable work on model training; similar scale (public).2 of 4 capabilitiesShared product typeCompareSStepFunPrivate · $10BAgent Optimizer and Step Plan do comparable work on agentic coding; similar scale (private).1 of 4 capabilitiesShared product typeCompareAAnthropicPrivate · $965BAgent Optimizer and Claude Agent SDK do comparable work on agentic coding; similar scale (private).2 of 4 capabilitiesShared product typeCompareCCodeRabbitPrivate · $1.5BAgent Optimizer and CodeRabbit Agent for Slack do comparable work on agentic coding; similar scale (private).1 of 4 capabilitiesShared product typeCompareOOpenCodePrivateAgent Optimizer and OpenCode do comparable work on agentic coding; OpenCode's scale not recorded; ships AI agent rather than the same layer.1 of 4 capabilitiesDifferent layerComparePAPoolside AIPrivate · $620M raisedAgent Optimizer and poolside do comparable work on agentic coding; similar scale (private); ships AI agent rather than the same layer.1 of 4 capabilitiesDifferent layerCompare

Closest alternatives are pairs a vendor page itself compares. Similar products are matched from what each product does; nothing in that group is asserted as a rivalry.

Sources

Total funding

Company disclosure · 31 Aug 2026

Same page states "$70M in funding."

Employees

Company disclosure · 31 Aug 2026

Same page: "80 employees in 14 countries."

Founded

Company disclosure · 31 Aug 2026

Comet's own About page: "2017 Comet launched."

Open Comet ML in the directory

Something wrong here? Send a correction — quote this company id: comet-ml.