Pricing snapshot
LlamaParse is LlamaIndex’s commercial platform for parsing and extracting documents. Pay-as-you-go and Pro combine a monthly plan allowance with post-allowance credit charges; Enterprise uses custom pricing and a volume credit discount. Credit use varies by the selected parse tier and extraction options.
Choosing a tier
Use Pay-as-you-go to measure the credits your chosen parse tier and extraction options consume. Pro changes both the included monthly credit allowance and concurrent-job limits. Enterprise is the option for single-tenant deployment, your own cloud, on-premises deployment, Enterprise SSO, or HIPAA-compliant processing with a BAA.
Published plans
Scroll sideways to see all columns. With a keyboard, focus this table and use the arrow keys.
| Plan | Monthly price | Included credits | Credit price | Limits and support |
|---|---|---|---|---|
Pay-as-you-go | $0/month | 10,000 credits every month | Then $1.50 per 1,000 credits | 1 project; 100 users; Email support |
Pro | $500/month | 400,000 credits every month | Then $1.25 per 1,000 credits | 5 projects; 100 users; Priority Slack support |
Enterprise | Custom | Custom | Volume discount on credits | Enterprise SSO; single-tenant, own-cloud or on-premises deployment; dedicated account manager |
Show full plan details
Scroll sideways to see all columns. With a keyboard, focus this table and use the arrow keys.
| Listed item | Pay-as-you-go | Pro | Enterprise |
|---|---|---|---|
Concurrent parse jobs | 5 | 20 | 100 |
Concurrent extract jobs | 5 | 20 | 100 |
Concurrent classify jobs | 5 | 20 | 100 |
Concurrent split jobs | 5 | 20 | 100 |
- Vendor source · Observed · Stored official pricing capture for llamaindex.
The comparison lists 130+ file formats, 80+ languages, and Markdown, Plain Text, per-page JSON, XLSX, HTML Tables, and Annotated PDF output.
How charges work
LlamaParse presents four parse tiers: Fast, Cost-effective, Agentic, and Agentic Plus, at 1, 3, 10, and 25 credits per page. Layout extraction adds 3 credits per page on any parse tier. Extract has Cost Effective and Agentic tiers and adds its own per-page credits on top of parsing; extraction targets can be per document, per page, or per table row. The free plan is described as 10,000 credits per month, about 1,000 pages, with agentic OCR, defined-schema extraction, and end-to-end document-agent deployment.
Monthly credits are used first, then purchased credits. When the balance reaches zero, new jobs pause while already-running jobs finish. Monthly credits reset each month and unused monthly credits do not carry over. Re-parsing a cached file costs nothing.
Conditions worth checking
Pro bonus. The 800,000-credit ($1,000) bonus is one-time, limited-time, restricted to business email domains, and unused bonus credits expire immediately on downgrade.
Deployment. Managed SaaS is available in the US or EU. Single-tenant SaaS, deployment in your own AWS, Azure, or Google Cloud environment, and on-premises deployment are Enterprise options. LlamaParse is also available on the AWS and Microsoft Azure marketplaces.
Data and compliance. Managed SaaS data is encrypted in transit and at rest; cached results are deleted after 48 hours and caching can be turned off. The vendor states a SOC 2 Type II report, GDPR support through an EU region and DPA, and HIPAA-compliant Enterprise processing with a BAA on request. The plan comparison lists 99.9% uptime.
Open-source scope. LlamaParse is commercial. LiteParse is a local open-source parser, and the separate LlamaIndex framework is open source under the MIT license; LlamaParse can be used with that framework or standalone.
Startups. LlamaIndex invites startups to apply for free credits, dedicated support, community spotlights, and more.
Our take
If your document mix is not yet measured, begin with the monthly Pay-as-you-go credits and test the intended parse and extraction settings. Move to Pro when its monthly credit allowance or higher concurrent-job limits match a measured workload. Use Enterprise when deployment, SSO, HIPAA processing with a BAA, or 100 concurrent jobs is required.