---
type: "post"
id: "introducing-extract-v2-5-analysis-ff7860"
slug: "introducing-extract-v2-5-analysis-ff7860"
title: "LlamaIndex announced Extract v2.5, an update to its schema-based document extraction agents"
description: "Extract v2.5 is presented as a shift from generic document reading toward schema-aware agents that allocate reasoning and verification effort to document structure."
retrieval_nugget: "Extract v2.5 is presented as a shift from generic document reading toward schema-aware agents that allocate reasoning and verification effort to document structure."
published_at: "2026-10-01"
updated_at: "2026-10-01"
record_date: "2026-10-01"
date_kind: "published_at"
topics: []
source_urls: ["https://www.llamaindex.ai/blog/introducing-extract-v2-5"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:introducing-extract-v2-5-analysis-ff7860"
status: "published"
visuals: []
editorial_provenance: {"schema_version":"newruntime-editorial-copy-v1","content_status":"source_grounded_final","final_copy_sha256":"sha256:291453256ef2a71ed07785f4028f70d531b92a13ae36eec06131c10c3d824cb8","reviewed_at":"2026-10-01T20:05:19.305Z","source_evidence_count":1,"verified_claim_count":3,"site_analysis_schema_version":"newruntime-site-analysis-v1","site_object_kind":"field_note","observed_fact_count":3,"implication_count":1,"watch_condition_count":1,"related_record_count":0}
analysis: {"schema_version":"newruntime-site-analysis-v1","object_kind":"field_note","thesis":"the release’s central change is combining higher reported extraction accuracy with stronger source-level grounding, rather than improving text capture alone.","observed_facts":[{"text":"LlamaIndex reports ExtractBench value F1 rising from 87.1 to 93.9 for Cost Effective, 89.8 to 95.8 for Agentic, and 95.1 to 96.4 for Agentic Plus, with no per-page price increase.","source_urls":["https://www.llamaindex.ai/blog/introducing-extract-v2-5"]},{"text":"It also says grounding scores rose from 46.8 to 80.6 for Agentic and 46.4 to 82.2 for Agentic Plus.","source_urls":["https://www.llamaindex.ai/blog/introducing-extract-v2-5"]},{"text":"In a 17-page filing example, the company says Cost Effective v2.5 returned all 238 holdings, compared with 87 under v2.0.","source_urls":["https://www.llamaindex.ai/blog/introducing-extract-v2-5"]}],"mechanism":"LlamaIndex says it introduced a document-extraction agent harness using structural reasoning, intermediate representations, purpose-built tools, and adaptive inspection; for Agentic and Agentic Plus, Advanced Citations return bounding boxes for evidence and confidence scores can help flag values for review.","why_now":"document extraction is often used to feed downstream databases and review workflows, where an omitted record or an unsupported value can matter as much as a transcription error; the update explicitly targets those failure modes.","implications":["if the reported gains hold on customer workloads, teams may be able to automate more schema-bound extraction while preserving a human verification path for uncertain fields."],"evidence_boundary":"all scores and examples here come from LlamaIndex’s product announcement and its ExtractBench framing; no independent replication, dataset details sufficient to generalize, or customer-side validation is supplied in the source.","watch_conditions":["independent tests on varied real-world documents should compare both field accuracy and citation quality at equivalent per-page costs before users treat the published gains as general performance guarantees."],"related_records":[],"new_branch_reason":"No prior New Runtime record was supplied, so this release starts a distinct analytical record focused on the version’s reported accuracy and grounding changes."}
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/introducing-extract-v2-5-analysis-ff7860/","markdown":"https://newruntime.com/posts/introducing-extract-v2-5-analysis-ff7860.md","json":"https://newruntime.com/posts/introducing-extract-v2-5-analysis-ff7860.json"}
---

# LlamaIndex announced Extract v2.5, an update to its schema-based document extraction agents

## Retrieval answer

Extract v2.5 is presented as a shift from generic document reading toward schema-aware agents that allocate reasoning and verification effort to document structure.

Extract v2.5 is presented as a shift from generic document reading toward schema-aware agents that allocate reasoning and verification effort to document structure. **Thesis:** the release’s central change is combining higher reported extraction accuracy with stronger source-level grounding, rather than improving text capture alone. **Mechanism:** LlamaIndex says it introduced a document-extraction agent harness using structural reasoning, intermediate representations, purpose-built tools, and adaptive inspection; for Agentic and Agentic Plus, Advanced Citations return bounding boxes for evidence and confidence scores can help flag values for review.

**Observed facts:** LlamaIndex reports ExtractBench value F1 rising from 87.1 to 93.9 for Cost Effective, 89.8 to 95.8 for Agentic, and 95.1 to 96.4 for Agentic Plus, with no per-page price increase. It also says grounding scores rose from 46.8 to 80.6 for Agentic and 46.4 to 82.2 for Agentic Plus. In a 17-page filing example, the company says Cost Effective v2.5 returned all 238 holdings, compared with 87 under v2.0. **Why now:** document extraction is often used to feed downstream databases and review workflows, where an omitted record or an unsupported value can matter as much as a transcription error; the update explicitly targets those failure modes.

**First implication:** if the reported gains hold on customer workloads, teams may be able to automate more schema-bound extraction while preserving a human verification path for uncertain fields. **Evidence boundary:** all scores and examples here come from LlamaIndex’s product announcement and its ExtractBench framing; no independent replication, dataset details sufficient to generalize, or customer-side validation is supplied in the source. **First watch condition:** independent tests on varied real-world documents should compare both field accuracy and citation quality at equivalent per-page costs before users treat the published gains as general performance guarantees.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)