> ## Documentation Index > Fetch the complete documentation index at: https://docs.aireserve.com/llms.txt > Use this file to discover all available pages before exploring further. # List GPU offers > What AI Reserve can provide right now, one row per accelerator at the requested GPU count, folded across every backing provider: best stock, lowest in-stock running price, regions with stock, and the capabilities at least one in-stock location supports. A capability that is false here (for example `persistentVolumes`) cannot be requested for that GPU right now, so clients grey the control instead of letting a launch fail. Nothing is reserved and no provider is named. Requires compute `read` access. ## OpenAPI ````yaml /api-reference/openapi.json get /v1/gpu/offers openapi: 3.0.3 info: title: AI Reserve Gateway API description: >- The AI Reserve API surface — OpenAI-compatible chat completions, Anthropic-compatible messages, Bedrock Converse, images, video, files, text-to-speech, realtime voice sessions, provider-hidden GPU infrastructure, and fine-tuning — all behind one endpoint and one API key. version: 1.0.0 servers: - url: https://api.aireserve.com security: - bearerAuth: [] tags: - name: Customers description: >- Per-end-customer usage, cost, and billable amounts for organizations that tag requests with their own customers. - name: GPU Infrastructure description: >- Provision and manage provider-hidden GPU instances. The recommended Instances API accepts workload requirements and an hourly price ceiling; AI Reserve handles provider placement internally. Every mutation is asynchronous and returns an operation to poll. - name: Customer keys description: >- Mint, list, and revoke API keys pinned to your own end customers (and their organizations), so spend splits per customer without per-request tagging. - name: AWS CLI login description: >- Browser-authorization endpoints consumed by `aws login --profile …` (AWS CLI v2.32.0+). All three routes are unauthenticated — the caller has no credential until `/v1/token` succeeds. - name: Fine-tuning description: >- Fine-tune open models on your own data. Data lives in an S3 bucket inside your organization's AWS hub account; you name a base, LoRA or full, a price ceiling and how the result must be servable, and AI Reserve places the run on a backend internally. The result is a model in your bucket you can deploy or download. paths: /v1/gpu/offers: get: tags: - GPU Infrastructure summary: List GPU offers description: >- What AI Reserve can provide right now, one row per accelerator at the requested GPU count, folded across every backing provider: best stock, lowest in-stock running price, regions with stock, and the capabilities at least one in-stock location supports. A capability that is false here (for example `persistentVolumes`) cannot be requested for that GPU right now, so clients grey the control instead of letting a launch fail. Nothing is reserved and no provider is named. Requires compute `read` access. operationId: listGpuOffers parameters: - name: count in: query required: false description: GPUs per instance (default 1). Stock is read per count. schema: type: integer minimum: 1 maximum: 16 - name: region in: query required: false description: Only stock in this normalized region. schema: type: string enum: - north_america - united_states - canada - europe - asia_pacific - oceania - name: accelerator in: query required: false description: Only this accelerator; default every accelerator. schema: type: string enum: - A40_48GB - A100_80GB - H100_80GB - H100_NVL_94GB - H200_141GB - B200_180GB - B300_288GB - L40_48GB - L40S_48GB - RTX_2000_ADA_16GB - RTX_4000_ADA_20GB - RTX_6000_ADA_48GB - RTX_A5000_24GB - RTX_PRO_4000_24GB - RTX_PRO_4500_32GB - RTX_PRO_6000_96GB - RTX_PRO_6000_MIG_24GB - RTX_PRO_6000_MIG_48GB - MI300X_192GB - name: diskGb in: query required: false description: Root disk the price includes (default 50). schema: type: integer minimum: -9007199254740991 maximum: 9007199254740991 - name: persistentGb in: query required: false description: >- Size of a persistent volume the launch would create; `persistentVolumes` then reflects where one of that size can be hosted and priced. schema: type: integer minimum: -9007199254740991 maximum: 9007199254740991 responses: '200': description: Priced. The request can be launched as quoted. content: application/json: schema: type: object properties: data: type: array items: type: object properties: accelerator: type: string description: >- Provider-independent accelerator model, e.g. `H100_80GB`. Pass as `accelerator` to quote or launch. label: type: string memoryGb: type: integer minimum: -9007199254740991 maximum: 9007199254740991 description: Memory per GPU in GB. manufacturer: type: string enum: - NVIDIA - AMD count: type: integer minimum: -9007199254740991 maximum: 9007199254740991 description: GPUs per instance this offer was read for. availability: type: string enum: - NONE - LOW - MEDIUM - HIGH - UNKNOWN description: >- Best stock across every eligible location. `NONE` means the platform cannot provide this right now. fromUsdPerHour: nullable: true description: >- Lowest in-stock running price (GPU + root disk) at this count; null when nothing is in stock. type: string pattern: ^\d+(?:\.\d{1,6})?$ regions: type: array items: type: string description: >- Normalized regions with stock. Empty when stock is in a location without a public region. capabilities: type: object properties: start: type: boolean stop: type: boolean restart: type: boolean ssh: type: boolean logs: type: boolean persistentVolumes: type: boolean stopPreservesRootDisk: type: boolean required: - start - stop - restart - ssh - logs - persistentVolumes - stopPreservesRootDisk additionalProperties: false description: >- What at least one in-stock location can do. A client greys a control exactly when its capability is false here: `persistentVolumes` false means no persistent storage can be created or mounted with this GPU right now. required: - accelerator - label - memoryGb - manufacturer - count - availability - fromUsdPerHour - regions - capabilities additionalProperties: false description: >- What AI Reserve can provide for one accelerator at one count, folded across every backing provider. Never names a provider. nextCursor: description: Pass as `?cursor=` to fetch the next page. type: string required: - data additionalProperties: false '400': description: Invalid request, missing Idempotency-Key, or schema violation. content: application/json: schema: $ref: '#/components/schemas/ComputeError' '401': description: Missing or invalid API key. content: application/json: schema: $ref: '#/components/schemas/ComputeError' '402': description: >- Wallet not enabled or not funded (`insufficient_funds`, `storage_delinquent`). content: application/json: schema: $ref: '#/components/schemas/ComputeError' '403': description: Key lacks the required compute access (`read` or `manage`). content: application/json: schema: $ref: '#/components/schemas/ComputeError' '404': description: Resource not found or owned by another organization. content: application/json: schema: $ref: '#/components/schemas/ComputeError' '409': description: >- `no_capacity` with `details` (see `ComputeNoCapacityDetails`): why nothing can be placed and what would work; or `volume_in_use` when the named volume is not ready or already attached. content: application/json: schema: $ref: '#/components/schemas/ComputeError' '503': description: Compute disabled or provider unavailable. content: application/json: schema: $ref: '#/components/schemas/ComputeError' security: - bearerAuth: [] - apiKeyAuth: [] components: schemas: ComputeError: type: object properties: error: type: object properties: message: type: string type: type: string enum: - invalid_request_error - authentication_error - permission_error - server_error param: nullable: true type: string code: type: string description: >- Machine-readable Compute error code, e.g. `quote_expired`, `price_changed`, `operation_conflict`, `no_capacity`, `volume_in_use`, `insufficient_funds`, `compute_access_required`, `not_found`. details: description: >- Machine-readable context for some codes. For `no_capacity` see `ComputeNoCapacityDetails`. type: object additionalProperties: {} required: - message - type - param - code additionalProperties: false required: - error additionalProperties: false description: OpenAI-shaped error envelope returned on every 4xx/5xx. securitySchemes: bearerAuth: type: http scheme: bearer description: >- Your AI Reserve API key (aireserve_api_…) from the API Keys page in the portal (https://portal.aireserve.com/keys). apiKeyAuth: type: apiKey in: header name: x-api-key description: The same API key sent as an `x-api-key` header. ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.