今天，我们推出 max effort，这是 Extract 的一个新设置，适用于agentic_plus层级。它旨在提供我们最高质量的抽取，并降低在你最关键、最复杂的文档上出现数据缺失或错误的风险。Max effort 根据所需计算量采用可变计费。

作为许多企业客户入驻流程的一部分，我们会开展概念验证，在他们计划用于生产的数据集样本上评估各种 Extract 配置。这一流程中的一个重要环节是找出既能满足其业务需求、又保持在预算内的最优层级和 schema。在许多情况下，比抽取服务价格更重要的成本，是某个值缺失或错误所带来的成本。客户会问：“我们能否多付一些钱来获得更高质量？”

在以下场景启用 max effort 时agentic_plus，答案现在是肯定的。我们开发了一些方法，能显著提升对一些最难文档的抽取效果。在 ExtractBench Pro 上，Value F1 从标准 Agentic Plus 的 77.7 提升到启用 max effort 的 92.9。启用 max effort 时，计费会根据抽取的复杂度和所需计算量而变化。这种额外的质量伴随着更多计算，成本可能高于标准agentic_plus任务。

你可以在下图中看到agentic_plus启用 max effort 如何扩展我们ExtractBench Pro成本与质量曲线的上端，为其最难的挑战提供另一种选择。

探索文档、schema 和答案键。

ExtractBench Pro · 100 份文档

Max effort 扩展了我们成本与质量曲线的上端

  • LlamaExtract
  • 编程智能体
  • 抽取 API
  • 30 分钟限制
  • 无限制

工作原理

例如，每页严格预算会限制抽取质量的一个方面是读取可视化图表。为了让 PDF 中的视觉内容可被模型读取，页面会被渲染成图像。为保持标准agentic_plus定价可预测，我们在图像分辨率上做出取舍，以适用于大多数文档。但在专门的图表类文档上，关键数据可能变得太小或模糊而无法读取。

With max effort, Extract can examine relevant regions at higher resolution when needed to pull out the pertinent information. The visualizations below show how that extra detail helps on some of our most visually complex documents.

The benefit of max effort is that it can apply more advanced extraction techniques as a document requires them. Standard tiers also adapt their work, but absorb that variability within a fixed price per page. With max effort, some documents can be pricier when additional work offers a large quality improvement, while in cases where standard methods are sufficient, the cost increase can be marginal.

Standard Agentic Plus → Max effort

Two different cost and quality tradeoffs

Compare all six categories
Standard Agentic Plus → Max effort. Value F1 uses a 30-minute limit.
CategoryValue F1Mean billed cost per document
Long dense pages12.7 → 98.4$0.075 → $3.82
Relational diagrams49.3 → 99.4$1.32 → $3.36
Document conventions97.5 → 99.5$2.98 → $4.29
Custom terminology98.6 → 99.0$3.25 → $3.33
Charts94.3 → 93.4$1.67 → $2.54
Counting67.6 → 58.3$1.41 → $1.50

If you would like to tryagentic_pluswith max effort, follow the guide below. We recommend running a representative sample of your documents with your existing schema before migrating to max effort. This will give you an idea of both the quality and cost you can expect with the configuration.

Use max when

  • Standardagentic_plusquality falls short
  • Your document is highly complex and dense
  • There are lots of visual elements such as charts or diagrams to interpret
  • Extraction errors are costly to your business

Extract V2.6 · Beta

Try max effort on your documents

Use Agentic Plus with effort set to max. Start with a small sample so you can compare quality and cost.

Python SDK

python -m pip install -U llama-cloud pydantic

The Python example takes your API key and a local PDF. Replace its sample schema with your own.

importos
importsys
frompathlibimportPath
fromllama_cloudimportLlamaCloud
frompydanticimportBaseModel

classInvoice(BaseModel):
total_amount:float

client=LlamaCloud(
api_key=os.environ["LLAMA_CLOUD_API_KEY"],
)
file=client.files.create(
file=Path(sys.argv[1]),
purpose="extract",
)
job=client.extract.run(
file_input=file.id,
configuration={
"tier":"agentic_plus",
"effort":"max",
"version":"latest",
"extraction_target":"per_doc",
"data_schema": Invoice.model_json_schema(),
"confidence_scores":True,
"cite_sources":True,
},
)
result=client.extract.get(
job.id,expand=["extract_metadata"],
)
print(result.extract_result)
print(result.extract_metadata)
配置指南 设置你的文档、schema 和 API key。 （在新标签页中打开）