> ## Documentation Index > Fetch the complete documentation index at: https://vastai-80aa3a82-auto-openapi-update-b65c39fe.mintlify.site/llms.txt > Use this file to discover all available pages before exploring further. # Get GPU Supply and Pricing History > Returns time-series supply/demand, pricing percentiles, and performance stats for one or more GPU types over a specified time range. ## OpenAPI ````yaml /api-reference/openapi.yaml get /api/v0/metrics/gpu/history/ openapi: 3.1.0 info: title: Vast.ai API description: >- Vast.ai REST API for managing GPU cloud instances, machine operations, and AI/ML workflows. ## AI Agent Quick-Start Install the CLI skill for your agent (Claude Code, Cursor, Windsurf, etc.): npx skills add vast-ai/vast-cli CLI reference: https://raw.githubusercontent.com/vast-ai/vast-cli/master/vastai/SKILL.md SDK reference: https://raw.githubusercontent.com/vast-ai/vast-cli/master/vastai_sdk/SKILL.md ## Auth All endpoints require `Authorization: Bearer $VAST_API_KEY`. Get your key at: https://cloud.vast.ai/manage-keys/ ## Key Quirks - `gpu_ram` in CLI = GB; in REST API = MB (CLI auto-converts) - SSH keys must be registered BEFORE creating an instance (VM: no recovery; Docker: can add post-create) - `onstart` field is limited to 4048 characters -- gzip+base64 for longer scripts - `POST /api/v0/asks/{id}/` (create instance) returns `new_contract` as the instance ID, not `id` - Poll trap: if `actual_status` becomes `exited`, `unknown`, or `offline` it will never reach `running` -- destroy and retry version: 1.0.0 contact: name: Vast.ai Support url: https://discord.gg/vast servers: - url: https://console.vast.ai description: Production server security: - BearerAuth: [] paths: /api/v0/metrics/gpu/history/: get: tags: - Machines summary: Get GPU Supply and Pricing History description: >- Returns time-series supply/demand, pricing percentiles, and performance stats for one or more GPU types over a specified time range. parameters: - name: gpu_name in: query required: false description: >- Comma-separated GPU names (e.g. "RTX 4090, H100_SXM") or "all". Defaults to RTX 5090, 4090, and 3090. schema: type: string example: RTX 4090 - name: verified in: query required: false description: Filter by verification status schema: type: string enum: - 'yes' - 'no' - all default: 'yes' - name: hosting_type in: query required: false description: Filter by hosting type schema: type: string enum: - all - secure_cloud - community default: all - name: num_gpus in: query required: false description: GPU count bucket or "all" schema: type: string default: all - name: start in: query required: false description: Unix timestamp for range start schema: type: integer - name: end in: query required: false description: Unix timestamp for range end schema: type: integer - name: step in: query required: false description: Seconds between data points (e.g. 3600 for hourly) schema: type: integer responses: '200': description: >- Success. Returns `{success: true, gpus: {: {supply_demand: {timestamps, rented_verified, avail_verified, rented_unverified, avail_unverified, total}, pricing: {timestamps, rented_p10, rented_median, rented_p90, avail_p10, avail_median, avail_p90}, stats: {tflops, dlperf, tflops_per_dollar, dlperf_per_dollar}}}}`. Returns `needs_machine: true` with empty gpus if host has no active machines. content: application/json: schema: type: object properties: success: type: boolean description: Always true on success needs_machine: type: boolean description: >- True when caller is a host with no active machines; gpus will be empty gpus: type: object description: Map of GPU name to its time-series data additionalProperties: type: object properties: supply_demand: type: object description: Time-series availability counts properties: timestamps: type: array items: type: integer description: Unix epoch timestamps for each data point bucket_size: type: integer description: >- GPUs per rig for bucketed queries; 1 when num_gpus=all total: type: array items: type: integer description: Total machine count per timestamp available: type: array items: type: integer description: >- Available machine count (alias for avail_verified+avail_unverified) rented_verified: type: array items: type: integer description: Rented verified machine count per timestamp avail_verified: type: array items: type: integer description: Available verified machine count per timestamp rented_unverified: type: array items: type: integer description: Rented unverified machine count per timestamp avail_unverified: type: array items: type: integer description: Available unverified machine count per timestamp total_gpus: type: array items: type: integer description: Raw total GPU count per timestamp rented_verified_gpus: type: array items: type: integer description: Raw rented verified GPU count per timestamp avail_verified_gpus: type: array items: type: integer description: Raw available verified GPU count per timestamp rented_unverified_gpus: type: array items: type: integer description: Raw rented unverified GPU count per timestamp avail_unverified_gpus: type: array items: type: integer description: Raw available unverified GPU count per timestamp pricing: type: object description: Time-series price percentiles per GPU-hour properties: timestamps: type: array items: type: integer description: Unix epoch timestamps rented_p10: type: array items: type: number description: 10th-percentile price for rented GPUs rented_median: type: array items: type: number description: Median price for rented GPUs rented_p90: type: array items: type: number description: 90th-percentile price for rented GPUs avail_p10: type: array items: type: number description: 10th-percentile price for available GPUs avail_median: type: array items: type: number description: Median price for available GPUs avail_p90: type: array items: type: number description: 90th-percentile price for available GPUs stats: type: object description: Latest performance statistics properties: tflops: type: number description: Latest 90th-percentile TFLOPS dlperf: type: number description: >- Latest 90th-percentile deep-learning performance score tflops_per_dollar: type: number description: >- TFLOPS per dollar at current median available price dlperf_per_dollar: type: number description: >- DL performance per dollar at current median available price '400': description: >- Invalid or missing parameters. Possible messages: "gpu_name is required", "verified must be 'yes', 'no', or 'all'", "hosting_type must be 'all', 'secure_cloud', or 'community'", "start, end, and step must be integers", "start must be <= end" content: application/json: schema: $ref: '#/components/schemas/Error' '403': description: 'Caller is not a host or admin: "Host or admin access required"' content: application/json: schema: $ref: '#/components/schemas/Error' security: - BearerAuth: [] components: schemas: Error: type: object properties: error: type: string msg: type: string securitySchemes: BearerAuth: type: http scheme: bearer description: API key must be provided in the Authorization header ````