Chonkie’s Chunking API is an enterprise-grade text processing service built on Chonkie OSS. Get instant access to all of our chunking algorithms, embedding generation, and text refinement. Avialble through simple REST APIs. No chunking logic to write, tokenizers to configure, or edge cases to debug. Just perfect chunks, delivered fast.
Key Features
All Our Chunkers
Use any of our awesome chunkers to split your data.
Each optimized for different document types and use cases.
Embeddings Refinery
Add vector embeddings to your chunks using any Hugging Face or OpenAI model.
Supports all major embedding providers.
Overlap Refinery
Add contextual overlap between chunks to prevent information loss at
boundaries. Configurable overlap sizes for optimal retrieval quality.
Multi-Language SDKs
Official Python and JavaScript/TypeScript SDKs with
full type safety and environment variable support.
Production-Ready
Battle-tested algorithms refined through thousands of real-world
deployments. Used by startups to enterprises for mission-critical RAG
systems.
Auto-Scaling
Automatically scales to handle your workload. From prototyping with single
documents to production pipelines processing millions of chunks.
Why Use Hosted Chunking?
Building chunking from scratch means dealing with:- Algorithm complexity: Implementing semantic chunking, code parsing, recursive splitting
- Tokenizer headaches: Managing multiple tokenizers, handling edge cases, counting accurately
- Performance optimization: Caching, batching, parallelization, memory management
- Maintenance burden: Updating dependencies, fixing bugs, handling new file formats
Instant Integration
Add chunking to your application in minutes with simple REST APIs or SDKs.
No complex setup or configuration required.
Battle-Tested Algorithms
Proven chunking strategies refined through thousands of production
deployments across diverse document types and use cases.
Always Up-to-Date
Automatic updates with new chunkers, optimizations, and bug fixes. You get
improvements without changing a line of code.
Use Cases
RAG Pipelines
RAG Pipelines
Preprocess documents for retrieval augmented generation. Chunk once, embed,
and store in your vector database for lightning-fast retrieval.
Document Analysis
Document Analysis
Break down large documents for analysis, summarization, or classification.
Process documents too large for single LLM calls.
Semantic Search
Semantic Search
Create search indices over documentation, support articles, or knowledge
bases. Combine with embeddings for powerful semantic search.
Code Indexing
Code Indexing
Index codebases for AI-powered code search, documentation generation, or
code review tools. Language-aware chunking preserves context.
Content Processing
Content Processing
Process blogs, articles, and content at scale. Perfect for content
recommendation systems or editorial tools.
Refineries: Enhance Your Chunks
Add Embeddings
Generate vector embeddings for your chunks using any Hugging Face model.
Supports all major providers: OpenAI, Cohere, Sentence Transformers, and
more.
Add Overlap
Add contextual overlap between chunks post-processing. Perfect for adding
overlap to existing chunks or experimenting with different overlap sizes.