Skip to content
LlamaIndex OSS Documentation
Guide
General
Pricing

Pricing

Detailed information about LlamaParse credit pricing and usage rates for parsing, indexing, extraction, classification, and split services.

All features are priced using credits, which are billed per page (or minute for audio). Credits vary by parsing mode, model, and whether files are cached.

Credit Rates

Section titled “Credit Rates”
RegionPrice per 1,000 Credits
North America$1.25
Europe$1.25

Parsing

Section titled “Parsing”

In the v2 API, parsing uses a simplified tier-based system. Choose a tier and the platform handles model selection automatically.

TierCredits per Page
Fast1
Cost-effective3
Agentic10
Agentic Plus45

Additional Configs:

  • Layout extraction: +3 credits per page (can be added to any tier)
  • Spreadsheet: 1 credit per sheet
  • Audio: 3 credits per minute

Indexing

Section titled “Indexing”
ModeCredits per Page (or Sheet)
Standard1
Spreadsheet2
Multi-modal2

Extraction

Section titled “Extraction”
ModeCredits per PageCredits per Page (extract only)*
Fast54
Balanced107
Multimodal2014
Premium6015

*For text files or pre-cached parsed files.

  • For text files (md, txt, csv, html), a page is defined as the equivalent of 600 tokens.
  • For text files or if the file has previously been parsed by LlamaParse, only extraction costs apply.
  • For Multimodal mode, 6 additional credits will be charged for docx/pptx format.

Split

Section titled “Split”
ModeCredits per Page
Default4 (3 for cached files)

If the file is already present in the LlamaParse cache, only split costs apply.

Classification

Section titled “Classification”
ModeCredits per Page
Fast1
Multimodal2

Sheets

Section titled “Sheets”

Sheets is currently in beta and free to use.

Agents

Section titled “Agents”

Agents is currently in beta and free to use. When agents make use of Parse, Extract, or Index, the pricing of the underlying module applies.

Cost Optimization Strategies

Section titled “Cost Optimization Strategies”

Parse

Section titled “Parse”
  1. Use caching — Parsed files are cached for 48 hours. Re-parsing the same file within that window is free.
  2. Choose the right tier — Start with Cost-effective (3 credits) for initial testing. Only move to Agentic (10) or Agentic Plus (45) when document complexity requires it. Fast (1 credit) outputs spatial text only — no markdown.
  3. Use page ranges — Parse only the pages you need instead of entire documents. Specify target_pages to avoid processing irrelevant content.

Cross-product

Section titled “Cross-product”
  1. Leverage extract-only pricing — If a file was previously parsed, subsequent extractions cost less (extract-only rates apply). Parse once, extract many times.
  2. Pre-filter with Classify — Use Classify (1-2 credits/page) to filter documents before running more expensive Parse or Extract jobs.
  3. Batch strategically — Group similar documents in a single job when possible rather than submitting many small requests.