# LlamaIndex — Open-source agent framework plus a credit-metered document parsing, extraction and indexing platform. > Source: The Agents Index — https://theagentsindex.com/llamaindex (structured, researched, re-verified) > Facts last verified: 2026-09-02 LlamaIndex is an open-source Python and TypeScript framework for building context-aware AI agents, paired with Workflows for multi-step orchestration. Its commercial LlamaParse platform sells agentic OCR, structured extraction, classification, splitting and managed indexes through one API. | Fact | Value | | --- | --- | | Website | https://www.llamaindex.ai | | Price | from $50/month | | Free tier | Yes | 10,000 credits per month (~1000 pages), 5 indexes, 50 files per index, file upload only as a data source, 20 requests/min, no pay-as-you-go: requests return 402 when credits run out ## What it does **Writes code** · Partly — Builder turns natural language into workflow code; it does not author application code from a description. > BuilderNatural language to code workflows builder > — https://www.llamaindex.ai/pricing · Compare plans table, Agents section, Builder row **Ships it** · Partly — What ships is the agent workflow itself: single-click deploy straight from a GitHub repo, listed on all four plans. > Single-click deploy of agentic workflows from GitHub repos > — https://www.llamaindex.ai/pricing · Compare plans table, Agents section, single-click deploy row **Done-for-you service** · No — You run the software: one API key, one SDK, five products; the highest-touch offering is a dedicated account manager. > LlamaParse is the enterprise platform for turning documents into production AI pipelines. One API key, one SDK, and five composable products > — https://developers.llamaindex.ai/llamaparse/ · Quickstart intro paragraph ## How much it does unattended **Runs autonomously** · Partly — Parse/extract/classify jobs run unattended via API, SDK or MCP; multi-step workflows are explicitly built to include human review. > Workflows is used to stitch together multiple steps, including combining human-in-the-loop review, document processing, and more, all in a single end-to-end workflow. > — https://www.llamaindex.ai/workflows · FAQ 02, What are common ways to use Workflows? **Multi-agent** · Yes — Workflows orchestrates multi-step systems and can combine several agents in one run. > Workflows allows developers to build highly controlled multi-step workflows that can even combine multiple agents. > — https://www.llamaindex.ai/workflows · FAQ 04, How is Workflows different from LlamaIndex? **Agent permissions** · Partly — Framework exposes human-in-the-loop and state as building blocks the developer wires up; the vendor publishes no policy engine that gates agent actions. > Leverage core components like memory, state management, human-in-the-loop, reflection, and more. > — https://www.llamaindex.ai/llamaindex · Extensible building blocks section, item 01 ## Which models it runs on **Claude** · Yes — Anthropic Claude is one of three model families the platform runs, via Anthropic API or AWS Bedrock. > Anthropic Claude | Anthropic API | AWS Bedrock > — https://developers.llamaindex.ai/llamaparse/self_hosting/ · LLM Provider Support table, Anthropic Claude row **GPT** · Yes — OpenAI models are supported directly or through Azure OpenAI. > OpenAI | OpenAI API | Azure OpenAI > — https://developers.llamaindex.ai/llamaparse/self_hosting/ · LLM Provider Support table, OpenAI row **Open models** · Partly — The open-source framework integrates HuggingFace and local models; the hosted platform's enumerated model families are OpenAI, Anthropic and Google only. > We support integrations with OpenAI, HuggingFace, Anthropic, and more. > — https://developers.llamaindex.ai/python/framework/module_guides/models/llms/ · Modules section **Model choice** · Partly — v1 API lets you name the model per page; the recommended v2 tier system picks the model for you. > In the v2 API, parsing uses a simplified tier-based system. Choose a tier and the platform handles model selection automatically. > — https://developers.llamaindex.ai/llamaparse/general/pricing/ · Parsing section, v2 (Tiers) tab ## Where you use it **In your editor** · Partly — Reached inside IDE agents through the hosted MCP server and coding-agent skills, not as a first-party IDE extension. > The LlamaParse MCP is a Model Context Protocol server that exposes the LlamaParse Platform's document-processing capabilities (Parse, Classify, Extract, Split, and Index) as tools that any MCP-compatible agent can call. > — https://developers.llamaindex.ai/llamaparse/for-agents/mcp/ · Intro paragraph **On the command line** · Yes — Skills and plugins install into terminal coding agents Claude Code and Codex; a Go CLI (llp) is a documented install path. > LlamaParse ships agent skills (focused capabilities that teach coding agents how to parse and extract content from documents) and agent plugins that bundle those skills (and, in one case, the LlamaParse MCP) for one-step installation in Claude Code and Codex. > — https://developers.llamaindex.ai/llamaparse/for-agents/skills/ · Intro paragraph **In your pipeline** · Partly — CI use is deployment only: a spec kept in source control, applied on every push. > Use llamactl deployments apply -f in GitHub Actions when you want a deployment spec in source control and an automated update on every push. > — https://developers.llamaindex.ai/python/llamaagents/llamactl/cd-with-github-actions/ · Continuous Deployment with GitHub Actions, opening line **In a browser** · Yes — Every product is usable from the hosted dashboard as well as SDK and REST API. > It's available as a web UI, Python SDK, and REST API. > — https://developers.llamaindex.ai/llamaparse/general/faq/ · Extract section, What is LlamaExtract? ## Whose machine it runs on **Self-hosted** · Partly — Full platform deploys into your own Kubernetes/VPC, but only on Enterprise plans. > Self-hosted LlamaCloud is available on Enterprise plans. > — https://developers.llamaindex.ai/llamaparse/self_hosting/ · Get Started section **Open source** · Partly — LlamaIndex and Workflows are open source; the LlamaParse commercial platform is not. > Our open source projects, LlamaIndex and Workflows, provide AI builders with the foundational building blocks for creating AI apps. > — https://www.llamaindex.ai/pricing · FAQ 01, Is LlamaParse Open Source? ## What it costs to run **How it meters** · Yes — Metered in credits billed per page (or per minute of audio) at $1.25 per 1,000 credits. > 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. > — https://developers.llamaindex.ai/llamaparse/general/pricing/ · Intro paragraph above Credit Rates **Free tier** · Yes — Free plan self-serve with 10,000 credits a month, no sales call. > It provides a generous provision of 10k free credits a month to all new users, but is not open source. > — https://www.llamaindex.ai/pricing · FAQ 01, Is LlamaParse Open Source? **API access** · Yes — One API key drives every product from Python, TypeScript, Go, Java or REST. > One API key, one SDK, and five composable products: Parse (agentic OCR), Extract (structured data), Classify, Split, and Index. > — https://developers.llamaindex.ai/llamaparse/ · Quickstart intro paragraph **MCP server** · Yes — Hosted MCP server with OAuth exposes parse, classify, extract, split and index tools to any MCP client, in NA and EU. > Instead of wiring up the REST API or SDK yourself, you connect your agent to the MCP server once and let it parse documents, extract structured data, classify and split files, and build, refresh, and retrieve from your indexes directly. > — https://developers.llamaindex.ai/llamaparse/for-agents/mcp/ · Intro paragraph **Bring your own key** · Yes — Self-hosted deployments call providers with your own keys; on the v1 hosted API 'Use your own API key' bills 1 credit per page. > All LLM calls go directly from your cluster to providers using your own API keys and contracts. > — https://developers.llamaindex.ai/llamaparse/self_hosting/ · LLM Provider Support section, closing paragraph ## Buying it for a team **A company can buy it** · Yes — Organizations hold members and projects; Enterprise is contracted with sales. > This is done through the use of Organizations, which has a list of team members that have access to all resources under that organization. > — https://developers.llamaindex.ai/llamaparse/general/organizations/ · Intro paragraph **Seat model** · No — Not sold per seat: Free, Starter and Pro each include 100 users and price on credits instead. > Users | 100 | 100 | 100 > — https://www.llamaindex.ai/pricing · Compare plans table, Platform section, Users row **Pooled budget** · Partly — Plan credits belong to the organization, but API keys and incurred billing stay scoped to the individual user. > Every user within an organization has the same level of access to the projects and other resources under that organization. However, API keys and any incurred billing remain scoped to each individual user. > — https://developers.llamaindex.ai/llamaparse/general/organizations/ · Closing paragraph **Admin controls** · No — Managed service has no role separation: organization members all hold the same access level. > Every user within an organization has the same level of access to the projects and other resources under that organization. > — https://developers.llamaindex.ai/llamaparse/general/organizations/ · Closing paragraph **Audit log** · No — The plan comparison's security rows stop at compliance, VPC, SSO/MFA and custom BAAs; no audit log is offered on any tier. > SOC2, HIPAA, and GDPR compliance > — https://www.llamaindex.ai/pricing · Compare plans table, Security & compliance section — its only rows are SaaS, SOC2/HIPAA/GDPR compliance, VPC, SSO & MFA and Custom BAAs **Single sign-on** · Partly — SSO & MFA is an Enterprise-only row; self-hosted deployments integrate OIDC with Entra ID or Okta. > Enterprise authentication - OIDC integration with your identity provider (Microsoft Entra ID, Okta, and others) > — https://developers.llamaindex.ai/llamaparse/self_hosting/ · Security and Compliance section, second bullet ## What happens to your code **Opt out of training** · Yes — Nothing is used for training by default, and do_not_cache=True skips the 48-hour cache entirely. > Your data is kept private to you only and is used only to return your results, never for model training. > — https://developers.llamaindex.ai/llamaparse/general/faq/ · Privacy & Caching, Are documents stored or cached? **Data residency** · Yes — Choose North America (us-east-1) or Europe (eu-central-1); uploaded data stays in its region, and regions cannot be migrated or linked. > Data will be stored within the region it is uploaded to. If interacting with LlamaCloud EU for example, all data provided will remain within the EU region for storage and processing. > — https://developers.llamaindex.ai/llamaparse/general/regions/ · Features section, Where is my data stored and processed? **Getting out** · Partly — Results export as markdown, JSON, XLSX or parquet and files can be deleted, but credits do not roll over and bonus credits vanish on downgrade. > Unused bonus credits expire immediately if you downgrade. > — https://www.llamaindex.ai/pricing · Pro plan card footnote **Certifications** · Yes — SOC 2 Type 2 certified, HIPAA compliant and GDPR-adherent; BAAs are Enterprise-only. > LlamaCloud adheres to the General Data Protection Regulation (GDPR) and all other applicable laws and regulations governing our services. We are also SOC 2 Type 2 certified and HIPAA compliant. > — https://developers.llamaindex.ai/llamaparse/general/regions/ · Legal & Compliance, privacy frameworks question ## Platform record (as of 2026-08-30) Reported by the code-hosting platform, not by the vendor. | Fact | Value | | --- | --- | | Licence | MIT | | Stars | 51919 | | Forks | 8054 | | Last push | 2026-08-29 | ## Provenance 1 of 23 sources behind the facts above are independent of the vendor. ## Coverage elsewhere None recorded. Every fact above comes from the vendor and is not independently corroborated yet. ## What changed - 2026-09-02 — Capabilities · Commercial terms · Free tier - 2026-08-31 — Capabilities · Commercial terms · Free tier · Price - 2026-08-23 — Alternatives · Documentation link · LinkedIn link · One-liner · Repository · Summary - 2026-08-17 — Tag:claude - 2026-08-16 — One-liner · Summary ## Correcting this record Anything marked "Not established" is a question we have not answered — it is not a No. If you represent LlamaIndex, or anything here is wrong or out of date, you can supply it with a source: the operations this site accepts from an agent are listed at https://theagentsindex.com/.well-known/agents.json — that catalogue needs no key to read, and each entry states its own auth. Supplying evidence is free. Nothing that can be bought here changes what a field says, whether a gap is closed, or how this record is scored.