Snowflake · Capability

Cortex Inference API — cortex-inference

Cortex Inference API — cortex-inference. 2 operations. Lead operation: Perform Llm Text Completion Inference.. Self-contained Naftiko capability covering one Snowflake business surface.

Run with Naftiko Snowflakecortex-inference

What You Can Do

POST
Cortexllminferencecomplete — Perform Llm Text Completion Inference.
/v1/api/v2/cortex/inference-complete
GET
Getmodels — Returns the Llms Available for the Current Session
/v1/api/v2/cortex/models

MCP Tools

perform-llm-text-completion-inference

Perform Llm Text Completion Inference.

returns-llms-available-current-session

Returns the Llms Available for the Current Session

read-only idempotent

Capability Spec

cortex-inference-cortex-inference.yaml Raw ↑
naftiko: 1.0.0-alpha2
info:
  label: Cortex Inference API — cortex-inference
  description: 'Cortex Inference API — cortex-inference. 2 operations. Lead operation: Perform Llm Text Completion Inference..
    Self-contained Naftiko capability covering one Snowflake business surface.'
  tags:
  - Snowflake
  - cortex-inference
  created: '2026-05-19'
  modified: '2026-05-19'
binds:
- namespace: env
  keys:
    SNOWFLAKE_API_KEY: SNOWFLAKE_API_KEY
capability:
  consumes:
  - type: http
    namespace: cortex-inference-cortex-inference
    baseUri: ''
    description: Cortex Inference API — cortex-inference business capability. Self-contained, no shared references.
    resources:
    - name: api-v2-cortex-inference:complete
      path: /api/v2/cortex/inference:complete
      operations:
      - name: cortexllminferencecomplete
        method: POST
        description: Perform Llm Text Completion Inference.
        outputRawFormat: json
        outputParameters:
        - name: result
          type: object
          value: $.
        inputParameters:
        - name: body
          in: body
          type: object
          description: Request body (JSON).
          required: false
    - name: api-v2-cortex-models
      path: /api/v2/cortex/models
      operations:
      - name: getmodels
        method: GET
        description: Returns the Llms Available for the Current Session
        outputRawFormat: json
        outputParameters:
        - name: result
          type: object
          value: $.
        inputParameters:
        - name: body
          in: body
          type: object
          description: Request body (JSON).
          required: false
  exposes:
  - type: rest
    namespace: cortex-inference-cortex-inference-rest
    port: 8080
    description: REST adapter for Cortex Inference API — cortex-inference. One Spectral-compliant resource per consumed operation,
      prefixed with /v1.
    resources:
    - path: /v1/api/v2/cortex/inference-complete
      name: api-v2-cortex-inference-complete
      description: REST surface for api-v2-cortex-inference:complete.
      operations:
      - method: POST
        name: cortexllminferencecomplete
        description: Perform Llm Text Completion Inference.
        call: cortex-inference-cortex-inference.cortexllminferencecomplete
        with:
          body: rest.body
        outputParameters:
        - type: object
          mapping: $.
    - path: /v1/api/v2/cortex/models
      name: api-v2-cortex-models
      description: REST surface for api-v2-cortex-models.
      operations:
      - method: GET
        name: getmodels
        description: Returns the Llms Available for the Current Session
        call: cortex-inference-cortex-inference.getmodels
        with:
          body: rest.body
        outputParameters:
        - type: object
          mapping: $.
  - type: mcp
    namespace: cortex-inference-cortex-inference-mcp
    port: 9090
    transport: http
    description: MCP adapter for Cortex Inference API — cortex-inference. One tool per consumed operation, routed inline through
      this capability's consumes block.
    tools:
    - name: perform-llm-text-completion-inference
      description: Perform Llm Text Completion Inference.
      hints:
        readOnly: false
        destructive: false
        idempotent: false
      call: cortex-inference-cortex-inference.cortexllminferencecomplete
      with:
        body: tools.body
      outputParameters:
      - type: object
        mapping: $.
    - name: returns-llms-available-current-session
      description: Returns the Llms Available for the Current Session
      hints:
        readOnly: true
        destructive: false
        idempotent: true
      call: cortex-inference-cortex-inference.getmodels
      with:
        body: tools.body
      outputParameters:
      - type: object
        mapping: $.