Skip to content
Z Zendikt
W
MLOps Platforms · Rank #1 of 10

Weights and Biases review and pricing

Largest neutral MLOps platform; CoreWeave acquired in May 2024.

By Weights and Biases (CoreWeave) · Founded 2017 · San Francisco, CA · private

Weights and Biases (W and B) is the largest neutral MLOps platform by installed base, founded 2017 by Lukas Biewald and Chris Van Pelt (both ex-CrowdFlower). The product covers experiment tracking (the original wedge), hyperparameter sweeps, artifact and dataset versioning, model registry, and a reports surface used by ML teams to share results across the organization. CoreWeave acquired W and B in May 2024 for a reported $1.7B, which is the single most consequential category event of the last 24 months. Post-acquisition product velocity has been mixed: customer-facing roadmap continues, but the long-standing concern is that CoreWeave (a GPU-cloud company) acquired a neutral cross-cloud MLOps platform, and ML teams running on AWS, Google Cloud, or Azure now have to weigh whether W and B remains genuinely neutral or quietly tilts toward CoreWeave compute. Strengths: deepest experiment-tracking surface, largest community and integration footprint, real model-registry and artifact-versioning, used at OpenAI, Anthropic, Nvidia, Toyota, and across most leading research labs. Trade-offs: per-user pricing scales aggressively at large teams, post-CoreWeave neutrality is a real question for multi-cloud buyers, the model-registry surface lags Vertex AI and SageMaker on enterprise governance, and self-hosted deployment is gated to the top tier.

Visit Weights and Biases Vendor pricing Compare to MLflow
Best for

ML engineering and research teams wanting the deepest neutral experiment tracking and model registry across PyTorch, TensorFlow, JAX, and Hugging Face. Particularly strong for research labs, foundation-model teams, and ML platform teams running multi-cloud or unwilling to commit to a single hyperscaler.

Worst for

Buyers already committed to one hyperscaler (Vertex AI, SageMaker, or Azure ML is usually cheaper and more integrated), buyers needing a strong feature store (Databricks Mosaic AI or SageMaker better), regulated buyers needing FedRAMP authorization (W and B is in-process at best), or buyers nervous about CoreWeave-related neutrality drift.

Vendor Trust Score

Is Weights and Biases a trustworthy vendor?

7.3/10
Mixed
Pricing transparency
Published rates; no hidden fees
7.5
Contract fairness
Reasonable terms; no auto-renew traps
7.5
Incident response
How they handle outages and breaches
8.0
Post-acquisition behavior
Customer treatment after M&A or PE
6.5
Executive stability
Leadership churn over 24 months
7.5
Roadmap honesty
Public commitments held
7.0
Trust signal log
  • 2024-05-14
    CoreWeave acquired Weights and Biases for a reported $1.7B
    Largest neutrality question in the MLOps category since the acquisition. CoreWeave is a GPU-cloud company; ML teams on AWS, GCP, Azure now weigh whether W and B remains genuinely neutral or quietly tilts toward CoreWeave compute over time.
  • 2024-11-12
    Roadmap continuity post-acquisition
    Customer-facing roadmap continues; some buyer reports of slower release cadence on non-acquisition-aligned features through late 2024 and into 2025.
  • 2025-04-22
    Renewal pricing crept up at large enterprises
    Several buyer reports of double-digit renewal increases through 2024 to 2025; consistent with broader post-acquisition pricing pattern.
Vendor Trust is scored independently of product quality. A great product from an unfair vendor still earns a low trust score.
Review Intelligence

What 520 reviews actually say

Synthesized from G2, Capterra, Reddit, Trustpilot. Patterns >15% prevalence shown.

Last synthesized
2026-04-29

Praise patterns

  • Deepest experiment-tracking surface in neutral MLOps
    87% →
  • Largest community and integration footprint
    78% →
  • Hyperparameter sweeps and artifact versioning in one product
    71% →
  • Reports surface useful for sharing ML results
    64% →

Complaint patterns

  • Per-user pricing scales aggressively at large teams
    51% ↑
  • CoreWeave acquisition raised neutrality concerns for multi-cloud buyers
    47% ↑
  • Self-hosted deployment gated to top tier
    41% →
  • Renewal pricing crept up post-acquisition
    38% ↑
Sentiment trend (6 months)
78/100 0 pts
12
01
02
03
04
05
Patterns are extracted from review corpus and human-verified. We surface trends, not anecdotes.
Verified Pricing

What buyers actually pay

312 anonymized deal disclosures · last updated 2026-05-01

Contribute your deal price
Company size Median annual
10 to 50 ML engineers (Pro) $6,000
50 to 500 ML engineers (Pro) $60,000
500+ ML engineers (Enterprise) $360,000
Verified pricing is crowdsourced from buyers under anonymity guarantees. Vendor-listed prices are validated against actual deals quarterly.
Compliance & Security

Auto-verified certifications

Verified 2026-05-01
SOC 2 Type II
ISO 27001
HIPAA
GDPR
CCPA
PCI DSS
◐ FedRAMP In-Process

Editorial: Strengths

  • Deepest experiment-tracking surface in the neutral MLOps category
  • Largest community footprint; default at OpenAI, Anthropic, Nvidia, Toyota
  • Hyperparameter sweeps, artifact versioning, model registry in one product
  • Strong integrations across PyTorch, TensorFlow, JAX, Hugging Face
  • Reports surface used to share ML results across the organization
  • Long-running stable APIs; portable across clouds and frameworks
  • CoreWeave acquisition gives multi-year capital runway

Editorial: Weaknesses

  • CoreWeave acquired W and B in May 2024 for a reported $1.7B; neutrality question for multi-cloud buyers
  • Per-user pricing scales aggressively at large ML teams
  • Model-registry governance lags Vertex AI and SageMaker on enterprise controls
  • Self-hosted deployment gated to the top tier (enterprise procurement burden)
  • Some buyer reports of slower roadmap velocity post-acquisition
  • Renewal pricing has crept up at large enterprises through 2024 to 2025
  • Feature store is thinner than SageMaker, Vertex, or Databricks Mosaic AI

Key features & integrations

  • +Experiment tracking across PyTorch, TensorFlow, JAX, Hugging Face
  • +Hyperparameter sweeps with Bayesian optimization
  • +Artifact and dataset versioning
  • +Model registry with stage promotion and lineage
  • +Reports surface for sharing ML results
  • +Tables and queries over experiment metadata
  • +Integrations with Slack, Jira, Linear, PagerDuty
  • +SAML SSO and audit log at Enterprise
  • +Self-hosted deployment at Enterprise
  • +REST API and Python SDK
80+ integrations
PyTorchTensorFlowJAXHugging FaceAWS SageMakerGoogle Vertex AIKubernetesSlackJiraGitHub
Geography supported
Global; strongest in US, EU, UK, India, Japan
Best fit
10 to 50,000 employees · ML engineering, research, and platform teams across solo researchers and large enterprises
Editorial deep-dive

Read our full ranking of MLOps Platforms

Weights and Biases ranks #1 in our editorial review of 10 mlops platforms platforms. The deep-dive covers methodology, comparison tables, decision matrix, migration scoring, and FAQs.

Read the full ranking

Closest alternatives in MLOps Platforms

#2
MLflow
Open-source MLOps baseline stewarded by Databricks.
★ 4.4 public
#3
Google Vertex AI
Google Cloud hyperscaler ML platform with deep BigQuery and Gemini integration.
★ 4.4 partial
#4
Amazon SageMaker
AWS hyperscaler ML platform with the broadest service breadth on the cloud.
★ 4.4 partial
Help the next buyer

Contribute your verified deal price

Pricing in B2B software is opaque because vendors want it that way. Verified buyer prices fix that, anonymously. Share what you actually paid for Weights and Biases; we’ll add it to the verified pricing dataset on this page (with company size band only, no identifying details).

Submit anonymously