> ## Documentation Index
> Fetch the complete documentation index at: https://docs.firecrawl.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# 搜索论文

可按主题、方法、基准、作者或类别搜索论文。结果包括规范 `paperId`、首选 `primaryId`、来源 ID、标题、摘要、评分，以及可选的排序信号。

有关工作流概览，请参见 [Research Index 指南](/zh/features/research)。


## OpenAPI

````yaml zh/api-reference/v2-openapi.json GET /search/research/papers
openapi: 3.0.0
info:
  contact:
    email: support@firecrawl.dev
    name: Firecrawl Support
    url: https://firecrawl.dev/support
  description: 用于与 Firecrawl 服务交互，执行网页抓取和爬取任务的 API。
  title: Firecrawl API
  version: v2
servers:
  - url: https://api.firecrawl.dev/v2
security:
  - bearerAuth: []
paths:
  /search/research/papers:
    get:
      tags:
        - Research
      summary: 搜索论文
      operationId: researchSearchPapers
      parameters:
        - description: 用于搜索论文的自然语言查询。
          in: query
          name: query
          required: true
          schema:
            minLength: 1
            type: string
        - description: 返回的排序论文的最大数量。
          in: query
          name: k
          required: false
          schema:
            default: 40
            maximum: 500
            minimum: 1
            type: integer
        - description: 作者子字符串过滤条件。可重复传入，或传入逗号分隔的值；所有过滤条件都必须匹配。
          in: query
          name: authors
          required: false
          schema:
            type: string
        - description: 论文类别过滤条件。可重复传入，或传入逗号分隔的值；所有过滤条件都必须匹配。
          in: query
          name: categories
          required: false
          schema:
            type: string
        - description: created/updated 日期的包含下限。
          in: query
          name: from
          required: false
          schema:
            format: date
            type: string
        - description: created/updated 日期的包含上限。
          in: query
          name: to
          required: false
          schema:
            format: date
            type: string
      responses:
        '200':
          content:
            application/json:
              example:
                results:
                  - abstract: >-
                      We show that diffusion models can achieve image sample
                      quality superior to the current state-of-the-art
                      generative models...
                    ids:
                      arxiv:
                        - '2105.05233'
                    paperId: '2014215642691656232'
                    primaryId: arxiv:2105.05233
                    score: 0.0163934
                    title: Diffusion Models Beat GANs on Image Synthesis
                success: true
              schema:
                $ref: '#/components/schemas/ResearchSearchPapersResponse'
          description: 排序后的论文结果。
        '400':
          description: 无效请求
        '401':
          description: 缺少 Bearer 令牌或令牌无效
        '429':
          description: 超出限流
        '500':
          description: 内部服务器错误
      security:
        - bearerAuth: []
components:
  schemas:
    ResearchSearchPapersResponse:
      properties:
        results:
          items:
            $ref: '#/components/schemas/ResearchPaperResult'
          type: array
        success:
          type: boolean
      required:
        - success
        - results
      type: object
    ResearchPaperResult:
      properties:
        abstract:
          type: string
        ids:
          $ref: '#/components/schemas/ResearchIdMap'
        paperId:
          description: 规范论文 ID，或对通过 SERP 发现的展示结果使用 web:<url>。
          type: string
        primaryId:
          description: 首选的 cite/fetch ID，例如 arxiv:<id>、pmid:<id>、pmcid:<id> 或 doi:<id>。
          type: string
        score:
          format: double
          type: number
        signals:
          $ref: '#/components/schemas/ResearchPaperSignals'
        title:
          type: string
      required:
        - paperId
        - primaryId
        - title
        - abstract
        - score
      type: object
    ResearchIdMap:
      additionalProperties:
        items:
          type: string
        type: array
      description: 按命名空间分组的来源标识符。
      example:
        arxiv:
          - '2105.05233'
      type: object
    ResearchPaperSignals:
      properties:
        articleRank:
          description: 结构扩展的 article-rank 分数。
          format: double
          type: number
        seedOverlap:
          description: 与该候选结果关联的不同种子数量。
          minimum: 0
          type: integer
        semantic:
          description: 意图搜索的语义分数。
          format: double
          type: number
        structural:
          description: 原始结构图信号。
          format: double
          type: number
      required:
        - structural
        - semantic
        - articleRank
        - seedOverlap
      type: object
  securitySchemes:
    bearerAuth:
      scheme: bearer
      type: http

````