What Explosion AI does

Explosion is a Berlin-based software company founded in October 2016 by Ines Montani (CEO) and Matthew Honnibal (CTO), the original authors of the spaCy library. The company builds developer tools for Artificial Intelligence, Machine Learning, and Natural Language Processing, with a portfolio anchored by three core products: spaCy, a free open-source NLP library downloaded 600M+ times and used by 139,000+ GitHub projects; Prodigy, a commercial annotation tool with 12,000+ users across 1,000+ companies that uses active learning to create training data for machine learning models; and Ellf, an AI-powered assistant for NLP development that integrates with coding agents like Claude Code (entered private beta in 2026). The technology stack is built in-house, including the Thinc deep learning library and curated-transformers framework, with a deliberate emphasis on modular, transparent, self-hostable tools that keep customer data on-premises.

The company's business model combines three revenue streams: commercial licensing of Prodigy on an annual subscription basis (pricing not publicly disclosed), custom NLP consulting and implementation services sold via upfront quoted engagements, and merchandise sales. Distribution is hybrid — open-source spaCy drives product-led growth and top-of-funnel awareness through GitHub, PyPI, blog, YouTube, conferences, and a free Advanced NLP course, while enterprise customers like Microsoft, Bayer, S&P Global, GitLab, and The Guardian are served through direct sales and tailored consulting engagements. Prodigy is sold via self-serve on its dedicated website prodi.gy with a free trial enabling self-evaluation before purchase.

Explosion bootstrapped for its first five years before raising a $6 million Series A from SignalFire in August 2021 at a $120 million post-money valuation, followed by a $3 million warrant exercise in July 2022. By 2024, the company returned to operating as a fully independent, self-sufficient entity with SignalFire retaining a 7.7% minority stake. Ownership remains concentrated with Honnibal (49.5%) and Montani (33%), alongside a 9.8% employee virtual share pool. The company has 11-50 employees and has stated it is not seeking further venture funding.

Name

Explosion AI

Legal name

Explosion

Website

https://explosion.ai

Company type

Private

Founded year

2016

Operating status

Operating

Headcount range

11–50 employees

Short description

Explosion AI is a Berlin-based software company founded in 2016 that builds developer tools for production NLP, including the open-source spaCy library (600M+ downloads), the commercial Prodigy annotation tool used by 1,000+ companies, and Ellf, an AI assistant for agentic NLP development, serving data scientists, ML engineers, and enterprises like Microsoft and Bayer.

Ownership category

akta.pro rank

Product category

Natural Language Processing Tools

NAICS

Software Publishers (5132), Custom Computer Programming Services (541511), Other Computer Related Services (541519), Computer Training (611420)

SIC

Services-Computer Programming Services (7371), Services-Prepackaged Software (7372), Services-Computer Integrated Systems Design (7373)

akta.pro primary industry

End-to-End Enterprise AI Platforms (MLOps & Model Lifecycle Management) (HDAEANAA)

akta.pro secondary industries

AI Application Enablement Platforms (Copilot/Agent Frameworks, SDKs) (HDAEANAJ), AI/ML Engineering & Model Development Services (BPAEAHAF), AI/ML Solution Integration & MLOps Enablement (BPAEAAAJ), Data Science, Machine Learning & AI (Modeling, MLOps, GenAI) (EDABAFAE)

Keywords

• Natural language processing
• Machine learning tools
• Data annotation software
• Developer AI libraries
• NLP consulting services

Headquarters

HQ city

Berlin

HQ country

Germany

HQ region

Europe

Offices1 record

Markets served

GTM type

B2B

Offering type

Software

Cost components

Personnel, Technology or R&D, Marketing or Sales, Operations, Others

Revenue model

1. Commercial Software Licensing (Prodigy): Prodigy is a commercial annotation tool sold as a subscription/license. It enables data scientists to annotate training data efficiently using active learning. The tool targets ML practitioners and data scientists who need to create high-quality training datasets.
2. NLP Consulting Services: Explosion offers custom NLP consulting and implementation services including data strategy, code review, NLP pipeline design, and annotation coaching. Services are quoted upfront with predictable fees, and typical response time is less than a week.
3. Merchandise Sales: The company sells branded merchandise including supporter packs, apparel, mugs, books, and accessories through its online store. Orders include free shipping worldwide with payments processed via Stripe or PayPal.
4. Open Source Community (Strategic): spaCy and other open-source tools serve as the primary community and awareness building channel. While free to use, they drive adoption and conversion to Prodigy and consulting services. The company does not require data to leave user servers.

Pricing tiers

Go-to-market motion3 records

Distribution channels4 records

Marketing channels8 records

Core offering

Explosion AI builds and sells developer tools for AI, machine learning, and natural language processing. Its core offerings are spaCy (a free open-source NLP library in Python), Prodigy (a commercial annotation tool for creating ML training data), Ellf (an AI-powered assistant for NLP development), and Thinc (a deep learning framework). The company also provides tailored NLP consulting and implementation services for enterprise clients.

Product overview

Explosion AI offers a portfolio of developer tools for AI and Natural Language Processing, comprising three core commercial products: spaCy (open-source NLP library), Prodigy (annotation tool for training data), and Ellf (AI-powered assistant for NLP development). The ecosystem includes numerous open-source libraries like spacy-llm (LLM integration), spacy-transformers, spacy-layout, and Thinc (deep learning library). These products work together as a cohesive stack: spaCy provides the production NLP foundation, Prodigy enables efficient training data creation, spacy-llm bridges LLMs with spaCy pipelines, and Ellf automates the development workflow. The company also offers tailored NLP consulting services.

Differentiator

Problem solved

Functional benefit

Brands

• spaCy: Free, open-source library for advanced Natural Language Processing in Python, designed for production use.
• Prodigy
• Ellf
• Thinc

Products and services

• spaCy Free, open-source library for advanced Natural Language Processing in Python, designed for production use. Supports information extraction, natural language understanding, and text preprocessing for deep learning, with features including named entity recognition, dependency parsing, part-of-speech tagging, tokenization, and text classification. Used by developers and data scientists building production NLP applications.
• Prodigy Modern commercial annotation tool for creating training data for machine learning models. Enables efficient data annotation for NER, intent detection, image classification, and other ML tasks with active learning capabilities. Targets ML practitioners and data scientists needing to create high-quality training datasets. Used by 12,000+ users across 1,000+ companies.
• Ellf AI-powered virtual assistant for NLP and machine learning projects that integrates with coding assistants like Claude Code for developing NLP solutions. Currently in private beta with workflow modules for annotation, auto-annotation, training experiments, and team collaboration.
• Thinc Lightweight open-source deep learning library powering spaCy, providing a flexible and configurable neural network framework usable independently for deep learning tasks.
• spacy-llm Open-source library integrating Large Language Models into structured spaCy NLP pipelines, enabling LLMs to work within spaCy's modular framework for reliable and efficient NLP solutions.
• spaCy Layout Open-source library that processes PDFs, Word documents, and other documents with spaCy, extending spaCy for document understanding beyond plain text.
• NLP Consulting Services Custom NLP consulting and implementation services including data strategy, code review, NLP pipeline design, and annotation coaching. Services are quoted upfront with predictable fees and typical response time is less than a week. Targets enterprises like Microsoft, Bayer, S&P Global, and GitLab needing custom NLP solutions.

Quantifiable outcome

• 600M+ downloads of spaCy library
• +4 more outcomes

Named customers6 records

Segments3 records

Ideal customer profiles3 records

Technology focussed Yes

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

Integration12 records

AI capability14 records

Feature7 records

Partnerships

Two partnerships are on record, tiered core and minor.

• Hugging FacecoreTechnology or Integration · 1 July 2021Partnership to integrate spaCy pipelines with the Hugging Face Hub. Users can upload any spaCy pipeline package to Hugging Face with a single command, including auto-generated model cards and metadata. This enables easy sharing of spaCy models within the broader ML ecosystem.
• META-NETminorStrategic or Co-development Partner · 1 January 2019Explosion was awarded the META Seal of Recognition at META-FORUM in Brussels, an international conference series backed by the European Union on language technologies for a multilingual information society.

Scale indicators11 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• Hugging Face: Hugging Face is the dominant open-source NLP/ML platform, offering the Transformers library and Hub — directly comparable to Explosion's spaCy + Hugging Face Hub integration partnership. Both serve the same developer persona (ML engineers building production NLP) and compete for community mindshare.
• Scale AI: Scale AI is a leading data annotation and labeling platform serving enterprise ML teams — directly competing with Prodigy in the data-preparation-for-ML category. Both target ML practitioners who need high-quality training data, though Scale is more service-heavy and Explosion is more software/product-led.
• Labelbox: Labelbox provides a data labeling and annotation platform for ML teams, competing directly with Prodigy. Both target enterprise data science and ML engineering teams building production training datasets, with overlapping use cases (NER, classification, image labeling).
• Snorkel AI: Snorkel AI is a data-centric AI platform that enables programmatic labeling and weak supervision — directly comparable to Prodigy's approach of accelerating training data creation. Both target enterprise ML teams seeking to scale annotation beyond manual labeling.

Emerging players

• John Snow Labs: John Snow Labs specializes in biomedical and clinical NLP, similar to how Explosion's spaCy is used in AstraZeneca's KAZU framework. Both serve regulated industries (healthcare/pharma) requiring production-grade, domain-specific NLP pipelines.
• Rasa: Rasa provides open-source conversational AI and NLP tooling for building chatbots — overlapping with Explosion's NLP developer-tools positioning. Both target developers building production NLP applications with an open-source-first model, though Rasa specializes in dialogue/intent while Explosion is broader NLP.

Broad incumbents

• Cohere: Cohere is an enterprise LLM/NLP platform offering cloud APIs for text understanding and generation — a broad incumbent whose cloud-based NLP services compete with Explosion's self-hosted, modular approach. Both target enterprise NLP use cases but diverge sharply on architecture (API vs. self-hosted).
• Databricks: Databricks offers a broad data + AI platform including MLflow for MLOps and integration with Hugging Face Transformers — comparable to Explosion's NLP-focused MLOps stack (Prodigy + spaCy + Thinc). Databricks is a much broader platform spanning data engineering, analytics, and ML.
• Amazon Web Services (Comprehend / SageMaker): AWS Comprehend and SageMaker provide managed NLP services and ML platforms, competing with Explosion's self-hosted tools. Both target enterprise ML practitioners, but AWS offers scalable cloud APIs while Explosion emphasizes data privacy and self-hosting.

Others

• Weights & Biases: Weights & Biases is an MLOps platform for experiment tracking, dataset versioning, and model evaluation — adjacent to Prodigy's evaluation plugins. Explosion already integrates with W&B, and both serve ML engineering teams building production models, though W&B is horizontal while Explosion is NLP-specialized.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat5 records

Key risks6 records

Key highlights7 records

Customer concentration

• GitHubgithub.com/explosion
• LinkedInlinkedin.com/company/explosion-ai
• Xx.com/explosion_ai

Revenue estimate

Valuation estimate

Number of profiles

Profiles2 records

Funding overview

Funding rounds1 record

Investors1 record

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Explosion AI