Parse documents into LLM-ready chunks

Scores as of 2026-09-14 (43 days old; Depot marks scores stale after 30 days). The pick is still Depot's current editorial pick; it has not been re-reviewed since. Prices below carry their own dates.

Depot Pick

LlamaParse · 8.22/10 for this job (curated estimate, high confidence)

Capability on tables/forms + dx + pricePerf of the cheap tier. Distinct from llm-ready-extract (web) and rag-memory (store).

Runner-up

Unstructured · 8.12/10

Scores higher than the pick on: Price / performance, Privacy / data retention, Lock-in risk (higher = worse).

When not to use the pick

Need connector-heavy ETL / VPC pipelines or OSS self-host partitioner — Unstructured; agentic forms/tables still fail LlamaParse bake-off — Reducto; you only have URLs not files — Firecrawl/Jina Reader.

Lock-in / migration

Runner-up Unstructured has lower stored lock-in risk (3.8 vs 4.8, 0–10, higher = more lock-in). Self-host fallback listed for this job: MarkItDown, Florence-2-base, PaddleOCR.

Sourced price (usd_per_1k_pages)

Pick: no sourced price in this job's unit (no_sourced_price).

Cheapest matching SKU: Reducto r-1 Parse API $10 — source, as of 2026-10-19. Not the pick; not quality-adjusted.

Price history: first seen 2026-10-19 · last changed no change recorded · history JSON

Free option

• Docling (oss) — source, as of 2026-10-25
• EasyOCR (oss) — source, as of 2026-10-25
• Florence-2-base (oss) — source, as of 2026-10-25

Failure modes to watch

• Collateral side effects: Snapshot the workspace or app state, then diff it against an allowlist. A passing check plus an unexpected diff is a failure.
• Hallucinated tool arguments: Reject a call that fails the current schema or names an unknown id, and do that before the tool runs a side effect.
• Partial tool success: Require a per-item status and a failure count. Success for the task requires the failed count to be zero or explicitly handed to a retry queue.
• Schema drift: Validate every tool call and every loaded table against the current schema. Quarantine mismatches instead of coercing them.
• Silent truncation: Tools must return whether the payload was truncated and how long the original was. Fail the task if the answer depends on the missing span.
• Stale retrieval: Store the source URL and the time it was fetched. If a newer canonical page exists, or two stored sources disagree, do not answer from the older one.

Reported outcomes (submitted reports only)

This job: loading… JSON · reports never change scores · seed/test records excluded

Depot Pick (LlamaParse) on this job: loading… JSON · reports never change scores · seed/test records excluded

Cite: Agent Provider Depot, Depot Pick for document-parse: llama-parse, scores as of 2026-09-14. Agents: GET /api/decide?jtbd=document-parse