LlamaIndex releases ExtractBench to evaluate 14 frontier systems
llama_index · x · 2026-08-26
LlamaIndex released ExtractBench to evaluate schema-guided extraction capabilities across 14 frontier systems. The benchmark covers 370 enterprise documents, 67 document types, and 4,800+ pages, targeting real-world challenges like scanned forms, nested tables, and complex financial reports. CTO Simon Suo will host a technical walkthrough discussing extraction difficulties, gaps in existing benchmarks, and the cost-accuracy trade-offs for VLMs, coding agents, and specialized APIs.
Related event: LlamaIndex Launches ExtractBench; Qwen Tops Document Extraction(3 posts)→
More from Models
- China's StepFun releases Step 5 Preview weights: 600B MoE, 27B active, 1M context — Dr_Singularity · 2026-10-15
- StepFun open-sources Step 5 Preview: 600B MoE with 1M context, No. 2 on OpenRouter — Dr_Singularity · 2026-10-15
- Google says teams across the company are using Gemini 4 Argon on its hardest engineering problems — CurieuxExplorer · 2026-10-15
- Perplexity's new embedding models: index with a 9B model, search with 0.6B — antoine_chaffin · 2026-10-15
- 797GB of n-gram counts in CPU RAM let 1B models match 57% more compute — PangWeiKoh · 2026-10-15
- Leak: Moonshot's Kimi flagship 'x4' done with pre-training, K3.1 in internal testing — lyraxana · 2026-10-15