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

# List Benchmarks

> Unified benchmark endpoint that aggregates scores from multiple benchmark sources (Artificial Analysis, Design Arena, and OpenRouter's own tau-bench, GPQA, and web-search evals). Filter by source to reproduce the exact shapes from the legacy per-source endpoints, or use task_type to find models suited for specific workloads. Artificial Analysis items carry one score per Artificial Analysis index (Intelligence Index, Finance & Accounting Index, Strategy & Ops Index, Legal Index, Healthcare & Medical Index, Engineering Index, Economics Index, Cyber Index, Openness Index, Multilingual Index, Decision Index); attribute Artificial Analysis when you republish them. A response whose items all come from one source carries that source's citation in meta; in a mixed response, attribute each item by its source. Use task_type=search (or a search_* benchmark_type) for OpenRouter's search benchmarks, which publish each model's highest-scoring eligible evaluation configuration with same-configuration runs combined by task-weighted mean. Authenticate with any valid OpenRouter API key. Rate-limited to 30 requests/minute per key and 500 requests/day per account.



## OpenAPI

````yaml /openapi/openapi.yaml get /benchmarks
openapi: 3.1.0
info:
  contact:
    email: support@openrouter.ai
    name: OpenRouter Support
    url: https://openrouter.ai/docs
  description: OpenAI-compatible API with additional OpenRouter features
  license:
    name: MIT
    url: https://opensource.org/licenses/MIT
  title: OpenRouter API
  version: 1.0.0
servers:
  - description: Production server
    url: https://openrouter.ai/api/v1
    x-speakeasy-server-id: production
security:
  - apiKey: []
tags:
  - description: API key management endpoints
    name: API Keys
  - description: Analytics and usage endpoints
    name: Analytics
  - description: Anthropic Messages endpoints
    name: Anthropic Messages
  - description: BYOK endpoints
    name: BYOK
  - description: >-
      Submit, list, poll, and delete asynchronous batches of inference requests.
      See https://openrouter.ai/docs/batch-quickstart.
    name: Batch
  - description: Benchmarks endpoints
    name: Benchmarks
  - description: Chat completion endpoints
    name: Chat
  - description: Task classification market-share endpoints
    name: Classifications
  - description: Containers endpoints
    name: Containers
  - description: Credit management endpoints
    name: Credits
  - description: >-
      Public OpenRouter usage datasets. Data returned by these endpoints is
      licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/):
      reuse and republish it, including commercially, with attribution to
      OpenRouter.
    name: Datasets
  - description: Text embedding endpoints
    name: Embeddings
  - description: Endpoint information
    name: Endpoints
  - description: External Users endpoints
    name: External Users
  - description: Files endpoints
    name: Files
  - description: Generation history endpoints
    name: Generations
  - description: Guardrail Policies endpoints
    name: Guardrail Policies
  - description: Guardrails endpoints
    name: Guardrails
  - description: Images endpoints
    name: Images
  - description: Model information endpoints
    name: Models
  - description: OAuth authentication endpoints
    name: OAuth
  - description: Observability endpoints
    name: Observability
  - description: Organization endpoints
    name: Organization
  - description: Presets endpoints
    name: Presets
  - description: Private Endpoints endpoints
    name: Private Endpoints
  - description: Private Providers endpoints
    name: Private Providers
  - description: Provider information endpoints
    name: Providers
  - description: Rerank endpoints
    name: Rerank
  - description: OpenAI-compatible Responses API endpoints
    name: Responses
  - description: >-
      Management endpoints for SCIM group-to-workspace mappings, authenticated
      with a management key. These are not the SCIM 2.0 connector endpoints for
      your identity provider. In your identity provider, enter the SCIM endpoint
      URL shown when you enable provisioning under Settings > Members > SCIM
      Mappings. See
      https://openrouter.ai/docs/guides/features/scim-mappings#set-up-provisioning.
    name: SCIM
  - description: Speech-to-text endpoints
    name: STT
    x-displayName: Transcriptions
  - description: >-
      System One endpoints for models such as Jev, compatible with the TypeSafe
      SDKs. See https://openrouter.ai/docs/guides/community/typesafe-sdk.
    name: SystemOne
    x-displayName: System One
  - description: Text-to-speech endpoints
    name: TTS
    x-displayName: Speech
  - description: >-
      The catalog of server tools OpenRouter runs on behalf of a model: accepted
      `tools[].type` spellings per API format, engines and pricing, and which
      endpoints run each tool natively. See
      https://openrouter.ai/docs/guides/features/server-tools.
    name: Tools
  - description: >-
      Store host-bound secrets for a workspace or for one intern. Scope is
      selected by the API key. Responses return metadata only, never secret
      values. See https://openrouter.ai/docs/guides/ori/vault.
    name: Vault
  - description: Video Generation endpoints
    name: Video Generation
  - description: Workspaces endpoints
    name: Workspaces
  - description: Alpha feature endpoints for Decisions requests
    name: alpha.decisions
  - description: Alpha feature endpoints for Extractions requests
    name: alpha.extractions
externalDocs:
  description: OpenRouter Documentation
  url: https://openrouter.ai/docs
paths:
  /benchmarks:
    get:
      tags:
        - Benchmarks
      summary: List Benchmarks
      description: >-
        Unified benchmark endpoint that aggregates scores from multiple
        benchmark sources (Artificial Analysis, Design Arena, and OpenRouter's
        own tau-bench, GPQA, and web-search evals). Filter by source to
        reproduce the exact shapes from the legacy per-source endpoints, or use
        task_type to find models suited for specific workloads. Artificial
        Analysis items carry one score per Artificial Analysis index
        (Intelligence Index, Finance & Accounting Index, Strategy & Ops Index,
        Legal Index, Healthcare & Medical Index, Engineering Index, Economics
        Index, Cyber Index, Openness Index, Multilingual Index, Decision Index);
        attribute Artificial Analysis when you republish them. A response whose
        items all come from one source carries that source's citation in meta;
        in a mixed response, attribute each item by its source. Use
        task_type=search (or a search_* benchmark_type) for OpenRouter's search
        benchmarks, which publish each model's highest-scoring eligible
        evaluation configuration with same-configuration runs combined by
        task-weighted mean. Authenticate with any valid OpenRouter API key.
        Rate-limited to 30 requests/minute per key and 500 requests/day per
        account.
      operationId: getBenchmarks
      parameters:
        - description: >-
            Benchmark source to query. Determines the shape of the returned
            items. When omitted, returns results from all sources.
          in: query
          name: source
          required: false
          schema:
            description: >-
              Benchmark source to query. Determines the shape of the returned
              items. When omitted, returns results from all sources.
            enum:
              - artificial-analysis
              - design-arena
              - openrouter
            example: artificial-analysis
            type: string
        - description: >-
            Filter results by task type. Every value except `search` names the
            Artificial Analysis index of the same name and returns only models
            scored on it, best first (`coding` and `agentic` are deprecated:
            Artificial Analysis has retired those indexes and returns no item
            for them). Design Arena answers only `coding`, `intelligence` and
            `agentic`, and OpenRouter's own evals only `intelligence` and
            `agentic`. `search` returns OpenRouter search benchmark results
            only.
          in: query
          name: task_type
          required: false
          schema:
            description: >-
              Filter results by task type. Every value except `search` names the
              Artificial Analysis index of the same name and returns only models
              scored on it, best first (`coding` and `agentic` are deprecated:
              Artificial Analysis has retired those indexes and returns no item
              for them). Design Arena answers only `coding`, `intelligence` and
              `agentic`, and OpenRouter's own evals only `intelligence` and
              `agentic`. `search` returns OpenRouter search benchmark results
              only.
            enum:
              - intelligence
              - coding
              - agentic
              - finance-and-accounting
              - strategy-and-ops
              - legal
              - healthcare-and-medical
              - engineering
              - economics
              - cyber
              - multilingual
              - decision
              - search
            example: intelligence
            type: string
        - description: >-
            Return results for one exact OpenRouter benchmark. A `search_*`
            value narrows the response to search results only; a classic value
            narrows the OpenRouter items and leaves other sources' items as they
            are.
          in: query
          name: benchmark_type
          required: false
          schema:
            description: >-
              Return results for one exact OpenRouter benchmark. A `search_*`
              value narrows the response to search results only; a classic value
              narrows the OpenRouter items and leaves other sources' items as
              they are.
            enum:
              - gpqa_diamond
              - tau_bench_verified_airline
              - search_browsecomp
              - search_hle
              - search_dsqa
              - search_widesearch
            example: search_widesearch
            type: string
        - description: >-
            Search benchmarks only: include the published lane configuration
            whitelist in each search item. Defaults to false. The whitelist is
            limited to agent turn count, reasoning effort, and temperature so
            future harness configuration changes do not change the public
            contract.
          in: query
          name: include_run_config
          required: false
          schema:
            default: false
            description: >-
              Search benchmarks only: include the published lane configuration
              whitelist in each search item. Defaults to false. The whitelist is
              limited to agent turn count, reasoning effort, and temperature so
              future harness configuration changes do not change the public
              contract.
            example: true
            type: boolean
        - description: 'OpenRouter search benchmarks only: filter by the search engine used.'
          in: query
          name: search_engine
          required: false
          schema:
            description: >-
              OpenRouter search benchmarks only: filter by the search engine
              used.
            example: exa
            type: string
        - description: >-
            OpenRouter search benchmarks only: filter by the request surface the
            lane ran on.
          in: query
          name: search_surface
          required: false
          schema:
            description: >-
              OpenRouter search benchmarks only: filter by the request surface
              the lane ran on.
            enum:
              - server-tool
              - plugin
            example: server-tool
            type: string
        - description: >-
            Design Arena only: arena to query. Defaults to `models` when source
            is `design-arena`.
          in: query
          name: arena
          required: false
          schema:
            description: >-
              Design Arena only: arena to query. Defaults to `models` when
              source is `design-arena`.
            enum:
              - models
              - builders
              - agents
            example: models
            type: string
        - description: >-
            Design Arena only: category within the arena (e.g. `codecategories`,
            `uicomponent`, `gamedev`, `3d`, `dataviz`, `image`, `video`, `svg`).
            When omitted, returns all categories.
          in: query
          name: category
          required: false
          schema:
            description: >-
              Design Arena only: category within the arena (e.g.
              `codecategories`, `uicomponent`, `gamedev`, `3d`, `dataviz`,
              `image`, `video`, `svg`). When omitted, returns all categories.
            example: codecategories
            type: string
        - description: >-
            Maximum number of items to return. When omitted, all matching
            results are returned.
          in: query
          name: max_results
          required: false
          schema:
            description: >-
              Maximum number of items to return. When omitted, all matching
              results are returned.
            example: 50
            minimum: 1
            type: integer
      responses:
        '200':
          content:
            application/json:
              example:
                data:
                  - agentic_index: null
                    coding_index: null
                    cyber_index: 36.4
                    decision_index: 75.7
                    display_name: GPT-4o
                    economics_index: 57.2
                    engineering_index: 60.4
                    finance_and_accounting_index: 58.1
                    healthcare_and_medical_index: 56.9
                    intelligence_index: 71.4
                    legal_index: 52.3
                    model_permaslug: openai/gpt-4o
                    multilingual_index: 0.82
                    openness_index: 33.3
                    pricing:
                      completion: '0.00001'
                      prompt: '0.0000025'
                    source: artificial-analysis
                    strategy_and_ops_index: 61.7
                  - accuracy: 0.72
                    accuracy_stddev: 0.03
                    avg_cost_per_task: 0.002
                    benchmark_type: gpqa_diamond
                    display_name: GPT-4o
                    last_run_timestamp: '2026-06-03T12:00:00Z'
                    model_permaslug: openai/gpt-4o
                    source: openrouter
                    total_tasks: 300
                meta:
                  as_of: '2026-06-03T12:00:00Z'
                  citation: null
                  model_count: 1
                  source: null
                  source_url: null
                  task_type: null
                  version: v1
              schema:
                $ref: '#/components/schemas/UnifiedBenchmarksResponse'
          description: >-
            Benchmark results filtered by the specified source and optional task
            type.
        '400':
          content:
            application/json:
              example:
                error:
                  code: 400
                  message: Invalid request parameters
              schema:
                $ref: '#/components/schemas/BadRequestResponse'
          description: Bad Request - Invalid request parameters or malformed input
        '401':
          content:
            application/json:
              example:
                error:
                  code: 401
                  message: Missing Authentication header
              schema:
                $ref: '#/components/schemas/UnauthorizedResponse'
          description: Unauthorized - Authentication required or invalid credentials
        '429':
          content:
            application/json:
              example:
                error:
                  code: 429
                  message: Rate limit exceeded
              schema:
                $ref: '#/components/schemas/TooManyRequestsResponse'
          description: Too Many Requests - Rate limit exceeded
        '500':
          content:
            application/json:
              example:
                error:
                  code: 500
                  message: Internal Server Error
              schema:
                $ref: '#/components/schemas/InternalServerResponse'
          description: Internal Server Error - Unexpected server error
components:
  schemas:
    UnifiedBenchmarksResponse:
      example:
        data:
          - agentic_index: null
            coding_index: null
            cyber_index: 36.4
            decision_index: 75.7
            display_name: GPT-4o
            economics_index: 57.2
            engineering_index: 60.4
            finance_and_accounting_index: 58.1
            healthcare_and_medical_index: 56.9
            intelligence_index: 71.4
            legal_index: 52.3
            model_permaslug: openai/gpt-4o
            multilingual_index: 0.82
            openness_index: 33.3
            pricing:
              completion: '0.00001'
              prompt: '0.0000025'
            source: artificial-analysis
            strategy_and_ops_index: 61.7
          - accuracy: 0.72
            accuracy_stddev: 0.03
            avg_cost_per_task: 0.002
            benchmark_type: gpqa_diamond
            display_name: GPT-4o
            last_run_timestamp: '2026-06-03T12:00:00Z'
            model_permaslug: openai/gpt-4o
            source: openrouter
            total_tasks: 300
        meta:
          as_of: '2026-06-03T12:00:00Z'
          citation: null
          model_count: 1
          source: null
          source_url: null
          task_type: null
          version: v1
      properties:
        data:
          items:
            oneOf:
              - $ref: '#/components/schemas/UnifiedBenchmarksAAItem'
              - $ref: '#/components/schemas/UnifiedBenchmarksDAItem'
              - $ref: '#/components/schemas/UnifiedBenchmarksORItem'
              - $ref: '#/components/schemas/UnifiedBenchmarksSearchItem'
          type: array
        meta:
          $ref: '#/components/schemas/UnifiedBenchmarksMeta'
      required:
        - data
        - meta
      type: object
    BadRequestResponse:
      description: Bad Request - Invalid request parameters or malformed input
      example:
        error:
          code: 400
          message: Invalid request parameters
      properties:
        error:
          $ref: '#/components/schemas/BadRequestResponseErrorData'
        openrouter_metadata:
          additionalProperties: {}
          type:
            - object
            - 'null'
        user_id:
          type:
            - string
            - 'null'
      required:
        - error
      type: object
    UnauthorizedResponse:
      description: Unauthorized - Authentication required or invalid credentials
      example:
        error:
          code: 401
          message: Missing Authentication header
      properties:
        error:
          $ref: '#/components/schemas/UnauthorizedResponseErrorData'
        openrouter_metadata:
          additionalProperties: {}
          type:
            - object
            - 'null'
        user_id:
          type:
            - string
            - 'null'
      required:
        - error
      type: object
    TooManyRequestsResponse:
      description: Too Many Requests - Rate limit exceeded
      example:
        error:
          code: 429
          message: Rate limit exceeded
      properties:
        error:
          $ref: '#/components/schemas/TooManyRequestsResponseErrorData'
        openrouter_metadata:
          additionalProperties: {}
          type:
            - object
            - 'null'
        user_id:
          type:
            - string
            - 'null'
      required:
        - error
      type: object
    InternalServerResponse:
      description: Internal Server Error - Unexpected server error
      example:
        error:
          code: 500
          message: Internal Server Error
      properties:
        error:
          $ref: '#/components/schemas/InternalServerResponseErrorData'
        openrouter_metadata:
          additionalProperties: {}
          type:
            - object
            - 'null'
        user_id:
          type:
            - string
            - 'null'
      required:
        - error
      type: object
    UnifiedBenchmarksAAItem:
      description: >-
        One model's Artificial Analysis index scores. Each score is the model's
        best on that index across the configurations Artificial Analysis
        evaluated (for example reasoning levels), so two indexes of one model
        can come from different configurations. The deprecated Coding Index and
        Agentic Index fields are null for every model: Artificial Analysis has
        retired those indexes.
      example:
        agentic_index: null
        coding_index: null
        cyber_index: 36.4
        decision_index: 75.7
        display_name: GPT-4o
        economics_index: 57.2
        engineering_index: 60.4
        finance_and_accounting_index: 58.1
        healthcare_and_medical_index: 56.9
        intelligence_index: 71.4
        legal_index: 52.3
        model_permaslug: openai/gpt-4o
        multilingual_index: 0.82
        openness_index: 33.3
        pricing:
          completion: '0.00001'
          prompt: '0.0000025'
        source: artificial-analysis
        strategy_and_ops_index: 61.7
      properties:
        agentic_index:
          deprecated: true
          description: >-
            Deprecated: Artificial Analysis has retired its Agentic Index and no
            longer publishes it. Always null.
          example: null
          format: double
          type:
            - number
            - 'null'
        coding_index:
          deprecated: true
          description: >-
            Deprecated: Artificial Analysis has retired its Coding Index and no
            longer publishes it. Always null.
          example: null
          format: double
          type:
            - number
            - 'null'
        cyber_index:
          description: >-
            Artificial Analysis Cyber Index score, on a 0-100 scale: the model's
            best across the configurations Artificial Analysis evaluated. Higher
            is better. Null when Artificial Analysis has not published a score
            for the model.
          example: 36.4
          format: double
          type:
            - number
            - 'null'
        decision_index:
          description: >-
            Artificial Analysis Decision Index score, on a 0-100 scale: the
            model's best across the configurations Artificial Analysis
            evaluated. Higher is better. Published for decision models and for
            the language models Artificial Analysis evaluates on decision tasks.
            Null when Artificial Analysis has not published a score for the
            model.
          example: 75.7
          format: double
          type:
            - number
            - 'null'
        display_name:
          description: Model name as listed on Artificial Analysis.
          example: GPT-4o
          type: string
        economics_index:
          description: >-
            Artificial Analysis Economics Index score, on a 0-100 scale: the
            model's best across the configurations Artificial Analysis
            evaluated. Higher is better. One of the six Artificial Analysis
            Capability Indexes. Null when Artificial Analysis has not published
            a score for the model.
          example: 57.2
          format: double
          type:
            - number
            - 'null'
        engineering_index:
          description: >-
            Artificial Analysis Engineering Index score, on a 0-100 scale: the
            model's best across the configurations Artificial Analysis
            evaluated. Higher is better. One of the six Artificial Analysis
            Capability Indexes. Null when Artificial Analysis has not published
            a score for the model.
          example: 60.4
          format: double
          type:
            - number
            - 'null'
        finance_and_accounting_index:
          description: >-
            Artificial Analysis Finance & Accounting Index score, on a 0-100
            scale: the model's best across the configurations Artificial
            Analysis evaluated. Higher is better. One of the six Artificial
            Analysis Capability Indexes. Null when Artificial Analysis has not
            published a score for the model.
          example: 58.1
          format: double
          type:
            - number
            - 'null'
        healthcare_and_medical_index:
          description: >-
            Artificial Analysis Healthcare & Medical Index score, on a 0-100
            scale: the model's best across the configurations Artificial
            Analysis evaluated. Higher is better. One of the six Artificial
            Analysis Capability Indexes. Null when Artificial Analysis has not
            published a score for the model.
          example: 56.9
          format: double
          type:
            - number
            - 'null'
        intelligence_index:
          description: >-
            Artificial Analysis Intelligence Index score, on a 0-100 scale: the
            model's best across the configurations Artificial Analysis
            evaluated. Higher is better. Null when Artificial Analysis has not
            published a score for the model.
          example: 71.4
          format: double
          type:
            - number
            - 'null'
        legal_index:
          description: >-
            Artificial Analysis Legal Index score, on a 0-100 scale: the model's
            best across the configurations Artificial Analysis evaluated. Higher
            is better. One of the six Artificial Analysis Capability Indexes.
            Null when Artificial Analysis has not published a score for the
            model.
          example: 52.3
          format: double
          type:
            - number
            - 'null'
        model_permaslug:
          description: Stable OpenRouter model identifier.
          example: openai/gpt-4o
          type: string
        multilingual_index:
          description: >-
            Artificial Analysis Multilingual Index score, as a fraction from 0
            to 1: the model's best across the configurations Artificial Analysis
            evaluated. Higher is better. Null when Artificial Analysis has not
            published a score for the model.
          example: 0.82
          format: double
          type:
            - number
            - 'null'
        openness_index:
          description: >-
            Artificial Analysis Openness Index score, on a 0-100 scale: the
            model's best across the configurations Artificial Analysis
            evaluated. Higher is better. Null when Artificial Analysis has not
            published a score for the model.
          example: 33.3
          format: double
          type:
            - number
            - 'null'
        pricing:
          $ref: '#/components/schemas/UnifiedBenchmarkPricing'
        source:
          description: Benchmark source discriminator.
          enum:
            - artificial-analysis
          type: string
        strategy_and_ops_index:
          description: >-
            Artificial Analysis Strategy & Ops Index score, on a 0-100 scale:
            the model's best across the configurations Artificial Analysis
            evaluated. Higher is better. One of the six Artificial Analysis
            Capability Indexes. Null when Artificial Analysis has not published
            a score for the model.
          example: 61.7
          format: double
          type:
            - number
            - 'null'
      required:
        - source
        - model_permaslug
        - display_name
        - intelligence_index
        - coding_index
        - agentic_index
        - finance_and_accounting_index
        - strategy_and_ops_index
        - legal_index
        - healthcare_and_medical_index
        - engineering_index
        - economics_index
        - cyber_index
        - openness_index
        - multilingual_index
        - decision_index
        - pricing
      type: object
    UnifiedBenchmarksDAItem:
      example:
        arena: models
        avg_generation_time_ms: 3200
        category: codecategories
        display_name: Claude Sonnet 4
        elo: 1423
        model_permaslug: anthropic/claude-sonnet-4
        pricing:
          completion: '0.000015'
          prompt: '0.000003'
        source: design-arena
        tournament_stats:
          first_place: 12
          fourth_place: 2
          second_place: 8
          third_place: 5
          total: 27
        win_rate: 72
      properties:
        arena:
          description: Arena this ranking belongs to.
          example: models
          type: string
        avg_generation_time_ms:
          description: Average generation time in milliseconds.
          example: 3200
          format: double
          type:
            - number
            - 'null'
        category:
          description: Category within the arena.
          example: codecategories
          type: string
        display_name:
          description: Human-readable model name from Design Arena.
          example: Claude Sonnet 4
          type: string
        elo:
          description: ELO rating from head-to-head arena battles.
          example: 1423
          format: double
          type: number
        model_permaslug:
          description: >-
            Stable OpenRouter model identifier when mapped; otherwise the
            upstream Design Arena model id.
          example: anthropic/claude-sonnet-4
          type: string
        pricing:
          $ref: '#/components/schemas/UnifiedBenchmarkPricing'
        source:
          description: Benchmark source discriminator.
          enum:
            - design-arena
          type: string
        tournament_stats:
          description: Placement distribution from tournament matches.
          properties:
            first_place:
              type:
                - integer
                - 'null'
            fourth_place:
              type:
                - integer
                - 'null'
            second_place:
              type:
                - integer
                - 'null'
            third_place:
              type:
                - integer
                - 'null'
            total:
              type:
                - integer
                - 'null'
          required:
            - first_place
            - second_place
            - third_place
            - fourth_place
            - total
          type: object
        win_rate:
          description: Win rate as a percentage (0–100).
          example: 72
          format: double
          type: number
      required:
        - source
        - model_permaslug
        - display_name
        - arena
        - category
        - elo
        - win_rate
        - avg_generation_time_ms
        - tournament_stats
        - pricing
      type: object
    UnifiedBenchmarksORItem:
      example:
        accuracy: 0.72
        accuracy_stddev: 0.03
        avg_cost_per_task: 0.002
        benchmark_type: gpqa_diamond
        display_name: GPT-4o
        last_run_timestamp: '2026-06-03T12:00:00Z'
        model_permaslug: openai/gpt-4o
        source: openrouter
        total_tasks: 300
      properties:
        accuracy:
          description: Aggregate accuracy score from 0 to 1. Higher is better.
          example: 0.72
          format: double
          type: number
        accuracy_stddev:
          description: Standard deviation of run accuracy, or null for a single run.
          example: 0.03
          format: double
          type:
            - number
            - 'null'
        avg_cost_per_task:
          description: Average cost per task in USD, or null if unavailable.
          example: 0.002
          format: double
          type:
            - number
            - 'null'
        benchmark_type:
          description: OpenRouter benchmark evaluation type.
          enum:
            - gpqa_diamond
            - tau_bench_verified_airline
          example: gpqa_diamond
          type: string
        display_name:
          description: Human-readable model name.
          example: GPT-4o
          type: string
        last_run_timestamp:
          description: Timestamp of the most recent public benchmark run.
          example: '2026-06-03T12:00:00Z'
          type: string
        model_permaslug:
          description: Stable OpenRouter model identifier.
          example: openai/gpt-4o
          type: string
        source:
          description: Benchmark source discriminator.
          enum:
            - openrouter
          type: string
        total_tasks:
          description: Total benchmark tasks across runs.
          example: 300
          type: integer
      required:
        - source
        - model_permaslug
        - display_name
        - benchmark_type
        - accuracy
        - accuracy_stddev
        - avg_cost_per_task
        - total_tasks
        - last_run_timestamp
      type: object
    UnifiedBenchmarksSearchItem:
      properties:
        avg_cost_per_task:
          description: Average cost per task in USD, or null if unavailable.
          example: 0.031
          format: double
          type:
            - number
            - 'null'
        avg_latency_per_task_ms:
          description: >-
            Average wall-clock latency per task in milliseconds, or null if
            unavailable.
          example: 45210
          format: double
          type:
            - number
            - 'null'
        benchmark_type:
          description: OpenRouter search benchmark.
          enum:
            - search_browsecomp
            - search_hle
            - search_dsqa
            - search_widesearch
          example: search_browsecomp
          type: string
        display_name:
          description: Human-readable model name.
          example: GPT-4o
          type: string
        last_run_timestamp:
          description: >-
            Timestamp of the newest qualifying run in the published
            configuration's lane.
          example: '2026-07-28T13:38:18Z'
          type: string
        model_permaslug:
          description: Stable OpenRouter model identifier.
          example: openai/gpt-4o
          type: string
        primary_metric:
          description: >-
            Identifies the meaning of `primary_score`: `f1_by_item` for
            WideSearch, `accuracy` for all other search benchmarks.
          enum:
            - accuracy
            - f1_by_item
          example: accuracy
          type: string
        primary_score:
          description: >-
            The benchmark's headline score from 0 to 1. Its meaning is
            identified by `primary_metric`: item-weighted F1 for WideSearch or
            strict accuracy for the other search benchmarks. Higher is better.
          example: 0.72
          format: double
          type: number
        run_config:
          $ref: '#/components/schemas/UnifiedBenchmarksSearchRunConfig'
        search_engine:
          description: Search engine the published configuration used.
          example: exa
          type: string
        search_surface:
          description: Request surface the published configuration went through.
          enum:
            - server-tool
            - plugin
          example: server-tool
          type: string
        source:
          description: Benchmark source discriminator.
          enum:
            - openrouter
          type: string
        total_tasks:
          description: Tasks evaluated across the published configuration's runs.
          example: 100
          type: integer
      required:
        - source
        - model_permaslug
        - display_name
        - benchmark_type
        - primary_metric
        - primary_score
        - total_tasks
        - avg_cost_per_task
        - avg_latency_per_task_ms
        - search_engine
        - search_surface
        - last_run_timestamp
      type: object
    UnifiedBenchmarksMeta:
      example:
        as_of: '2026-06-03T12:00:00Z'
        citation: >-
          Source: Artificial Analysis (artificialanalysis.ai) via OpenRouter
          (openrouter.ai/rankings).
        model_count: 50
        source: artificial-analysis
        source_url: https://artificialanalysis.ai
        task_type: null
        version: v1
      properties:
        as_of:
          description: ISO-8601 timestamp of when this data was last updated.
          example: '2026-06-03T12:00:00Z'
          type: string
        citation:
          description: >-
            Required attribution when republishing this data. Always set when
            `source` is set. When `source` is null, set if every returned item
            comes from one source, and null if the items span sources (attribute
            each item by its `source` discriminator) or there are none.
          example: >-
            Source: Artificial Analysis (artificialanalysis.ai) via OpenRouter
            (openrouter.ai/rankings).
          type:
            - string
            - 'null'
        model_count:
          description: Number of unique models in the response.
          type: integer
        source:
          description: >-
            The one source the response is limited to: the `source` filter, or
            Artificial Analysis for a `task_type` only it answers. Null when the
            response draws on every source.
          enum:
            - artificial-analysis
            - design-arena
            - openrouter
            - null
          example: artificial-analysis
          type:
            - string
            - 'null'
        source_url:
          description: >-
            URL of the upstream data source. Always set when `source` is set.
            When `source` is null, set if every returned item comes from one
            source, and null if the items span sources or there are none.
          example: https://artificialanalysis.ai
          type:
            - string
            - 'null'
        task_type:
          description: The task_type filter applied, or null if showing all.
          type:
            - string
            - 'null'
        version:
          description: Dataset version.
          enum:
            - v1
          type: string
      required:
        - as_of
        - version
        - source
        - source_url
        - citation
        - model_count
        - task_type
      type: object
    BadRequestResponseErrorData:
      description: Error data for BadRequestResponse
      example:
        code: 400
        message: Invalid request parameters
      properties:
        code:
          type: integer
        message:
          type: string
        metadata:
          additionalProperties: {}
          type:
            - object
            - 'null'
      required:
        - code
        - message
      type: object
    UnauthorizedResponseErrorData:
      description: Error data for UnauthorizedResponse
      example:
        code: 401
        message: Missing Authentication header
      properties:
        code:
          type: integer
        message:
          type: string
        metadata:
          additionalProperties: {}
          type:
            - object
            - 'null'
      required:
        - code
        - message
      type: object
    TooManyRequestsResponseErrorData:
      description: Error data for TooManyRequestsResponse
      example:
        code: 429
        message: Rate limit exceeded
      properties:
        code:
          type: integer
        message:
          type: string
        metadata:
          additionalProperties: {}
          type:
            - object
            - 'null'
      required:
        - code
        - message
      type: object
    InternalServerResponseErrorData:
      description: Error data for InternalServerResponse
      example:
        code: 500
        message: Internal Server Error
      properties:
        code:
          type: integer
        message:
          type: string
        metadata:
          additionalProperties: {}
          type:
            - object
            - 'null'
      required:
        - code
        - message
      type: object
    UnifiedBenchmarkPricing:
      description: >-
        OpenRouter pricing per token for this model. Null if pricing is
        unavailable.
      example:
        completion: '0.000015'
        prompt: '0.000003'
      properties:
        completion:
          description: Cost per output token (USD, decimal string).
          example: '0.000015'
          type: string
        prompt:
          description: Cost per input token (USD, decimal string).
          example: '0.000003'
          type: string
      required:
        - prompt
        - completion
      type:
        - object
        - 'null'
    UnifiedBenchmarksSearchRunConfig:
      description: >-
        Published lane configuration, included only when
        include_run_config=true. Only the agent turn count, reasoning effort,
        and temperature are exposed; other harness settings are intentionally
        not part of the public contract.
      properties:
        max_agent_turns:
          description: Agent-turn count for the published lane, or null for plugin lanes.
          example: 25
          type:
            - integer
            - 'null'
        reasoning_effort:
          description: >-
            Reasoning effort configured for the published lane, or null when
            omitted.
          example: high
          type:
            - string
            - 'null'
        temperature:
          description: >-
            Sampling temperature configured for the published lane, or null when
            omitted.
          example: 0.2
          format: double
          type:
            - number
            - 'null'
      required:
        - max_agent_turns
        - reasoning_effort
        - temperature
      type: object
  securitySchemes:
    apiKey:
      description: API key as bearer token in Authorization header
      scheme: bearer
      type: http

````