CHUNKER.CC is an AI-native document chunking API with micropayments on Solana. Upload documents and receive semantically chunked text optimized for RAG pipelines and LLM context windows.
https://api.adepti.cc
No authentication required. Simply POST your file to the endpoint. Limited to 100 pages/day.
Requires USDC payment on Solana. Payment flow:
X-PAYMENT headerX-PAYMENT: <solana-transaction-signature>Agent-friendly JSON endpoint. No multipart encoding - just send JSON with a document URL.
JSON body endpoint for AI agents. Accepts URL to document, returns chunks.
{
"url": "https://example.com/document.pdf",
"tier": "semantic",
"output_format": "pinecone"
}
| Field | Type | Description |
|---|---|---|
urlREQUIRED |
string | URL to fetch document from (PDF, TXT, HTML, DOCX) |
tierOPTIONAL |
string | "free", "structure" (default), "semantic", "ai", or "ai-pro" |
output_formatOPTIONAL |
string | "raw", "pinecone", "weaviate", "chroma", "qdrant" |
# 1. Request without payment - get 402 with payment requirements
curl -X POST https://api.adepti.cc/chunk \
-H "Content-Type: application/json" \
-d '{"url": "https://example.com/doc.pdf", "tier": "semantic"}'
# Response: 402 with paymentRequirements
# 2. Pay USDC to the specified address, get transaction signature
# 3. Retry with payment
curl -X POST https://api.adepti.cc/chunk \
-H "Content-Type: application/json" \
-H "X-PAYMENT: 5K4t...signature..." \
-d '{"url": "https://example.com/doc.pdf", "tier": "semantic"}'
Download results for large documents. When response has 100+ chunks, a download_url is returned instead of inline chunks.
# Large document response includes download_url
{
"success": true,
"total_chunks": 247,
"download_url": "https://api.adepti.cc/results/standard_a1b2c3d4e5f6",
"expires_in": "60 minutes"
}
# Fetch full results
curl https://api.adepti.cc/results/standard_a1b2c3d4e5f6
/chunk with JSON bodypayTo addressX-PAYMENT: {tx_signature} headerdownload_url separatelyReturns API info, pricing, and configuration.
curl https://api.adepti.cc/
Health check endpoint for monitoring.
curl https://api.adepti.cc/health
// Response
{ "status": "healthy", "network": "solana-mainnet" }
Basic recursive character chunking. Limited to 100 pages/day. Aliases: /chunk/free
| Parameter | Type | Description |
|---|---|---|
fileREQUIRED |
File | Document to chunk (.txt, .pdf) |
curl -X POST https://api.adepti.cc/chunk/demo \
-F "[email protected]"
Structure-aware chunking. Preserves code blocks, tables, lists. Aliases: /chunk/structure
| Parameter | Type | Description |
|---|---|---|
fileREQUIRED |
File | Document to chunk (.txt, .pdf) |
X-PAYMENTREQUIRED |
Header | Solana transaction signature |
curl -X POST https://api.adepti.cc/chunk/standard \
-H "X-PAYMENT: 5K4t...signature..." \
-F "[email protected]"
Embedding-based boundary detection. Splits where topics change using OpenAI embeddings.
| Parameter | Type | Description |
|---|---|---|
fileREQUIRED |
File | Document to chunk (.txt, .pdf) |
X-PAYMENTREQUIRED |
Header | Solana transaction signature |
curl -X POST https://api.adepti.cc/chunk/semantic \
-H "X-PAYMENT: 5K4t...signature..." \
-F "[email protected]"
LLM-powered chunking. Claude Haiku identifies natural semantic boundaries.
| Parameter | Type | Description |
|---|---|---|
fileREQUIRED |
File | Document to chunk (.txt, .pdf) |
X-PAYMENTREQUIRED |
Header | Solana transaction signature |
curl -X POST https://api.adepti.cc/chunk/ai \
-H "X-PAYMENT: 5K4t...signature..." \
-F "[email protected]"
Claude Sonnet for boundaries + LLM-generated summaries and entity extraction. Production RAG.
| Parameter | Type | Description |
|---|---|---|
fileREQUIRED |
File | Document to chunk (.txt, .pdf) |
X-PAYMENTREQUIRED |
Header | Solana transaction signature |
curl -X POST https://api.adepti.cc/chunk/ai-pro \
-H "X-PAYMENT: 5K4t...signature..." \
-F "[email protected]"
All chunking endpoints return the same response structure:
{
"success": true,
"tier": "standard",
"total_chunks": 12,
"metadata": {
"filename": "document.txt",
"content_type": "text/plain",
"size_bytes": 15420,
"estimated_pages": 3
},
"chunks": [
{
"text": "First chunk content...",
"metadata": {
"chunk_index": 0,
"char_count": 1024
}
},
{
"text": "Second chunk content...",
"metadata": {
"chunk_index": 1,
"char_count": 987
}
}
],
"payment": {
"amount": 1000,
"asset": "USDC",
"network": "solana-mainnet"
}
}
| Code | Status | Description |
|---|---|---|
200 |
Success | Request processed successfully |
400 |
Bad Request | Invalid file format, encoding, or corrupted PDF |
402 |
Payment Required | Missing or invalid payment for paid tier |
500 |
Server Error | Internal processing error |
{
"error": "Payment Required",
"paymentRequirements": {
"x402Version": 1,
"scheme": "exact",
"network": "solana-mainnet",
"maxAmountRequired": "4000",
"asset": "EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v",
"payTo": "8hTm...dWT",
"description": "Semantic tier embedding-based chunking"
}
}