Summary
Experiment tracker for foundation-model training, debugging, and training observability; acquisition by OpenAI announced.
Description
Neptune helps AI teams track experiments, monitor training, debug model behavior, and understand complex foundation-model training runs as they happen.
Positioning
Experiment tracker for foundation-model training
Key facts
- HQ location
- Warsaw, Poland / remote
- Founded
- 2017
- Employee range
- 51-200 (51-200 pre-acquisition)
- Funding stage
- Acquired
- Company type
- Unknown (Acquired / OpenAI announced)
- Pricing model
- Subscription (SaaS subscription; hosted service winding down post-acquisition)
- Last updated
Financials
- Revenue estimate
- Unknown
- Valuation estimate
- Deal terms undisclosed; Reuters reported under $400M stock value estimate
- Investments
- Acquisition by OpenAI announced Dec 2025; prior public funding was modest/undisclosed in detail
Relationships
- Target customers
- AI labs, ML researchers, and teams training large/foundation models
- Key competitors
- Weights & Biases, Comet, MLflow, WhyLabs, TensorBoard
- Known customers
- OpenAI, Samsung, Roche, HP and foundation-model teams publicly referenced
Classification (raw research text)
- Core focus
- Foundation-model experiment tracking
- Core industry
- AI Development Tools / MLOps
- Core category
- Experiment tracking and training observability
Shown verbatim from the research spreadsheet — deriving structured industry tags from this text is a future phase.
Segments, Industries & Certifications
- Segments