> ## Documentation Index > Fetch the complete documentation index at: https://docs.flex.ai/llms.txt > Use this file to discover all available pages before exploring further. # Usage and cost for a single request > Looks up one request's usage row for exact per-request cost attribution — the per-turn complement to the whole-key aggregate `/v1/flex/usage` returns. `request_id` is the value of the `x-litellm-call-id` response header on the original inference call (**not** the `id` field in the response body). Capture that header per call, then join it here. The lookup is scoped to the calling key: a request made with a different key returns the same `404` as an id that does not exist. Rows are written when a request completes, so a lookup immediately after the call may briefly `404` — retry after a few seconds. `total_spend` is computed through the same pricing path as the `/v1/flex/usage` aggregates, so per-request values sum to the totals you reconcile against. Only LiteLLM-routed (text) requests carry a call id, so rows always have `modality=text`. ## OpenAPI ````yaml /inference-api/openapi.yaml get /v1/flex/usage/request/{request_id} openapi: 3.1.0 info: title: FlexAI Inference API version: 1.0.0 description: | OpenAI-compatible inference API for text, code, reasoning, vision, and embedding models hosted by FlexAI. Use any OpenAI SDK by pointing the `base_url` at this service. contact: name: FlexAI Support email: support@flex.ai url: https://docs.flex.ai/inference-api/overview servers: - url: https://api.flex.ai description: Production - url: https://tokens.flex.ai description: Production (legacy host, still supported) security: - bearerAuth: [] tags: - name: Chat description: Chat completions (streaming, tool use, vision). - name: Completions description: Legacy text completions. - name: Embeddings description: Generate vector embeddings for text. - name: Models description: Discover available models. - name: Usage description: Programmatic spend and quota for the calling key. paths: /v1/flex/usage/request/{request_id}: get: tags: - Usage summary: Usage and cost for a single request description: | Looks up one request's usage row for exact per-request cost attribution — the per-turn complement to the whole-key aggregate `/v1/flex/usage` returns. `request_id` is the value of the `x-litellm-call-id` response header on the original inference call (**not** the `id` field in the response body). Capture that header per call, then join it here. The lookup is scoped to the calling key: a request made with a different key returns the same `404` as an id that does not exist. Rows are written when a request completes, so a lookup immediately after the call may briefly `404` — retry after a few seconds. `total_spend` is computed through the same pricing path as the `/v1/flex/usage` aggregates, so per-request values sum to the totals you reconcile against. Only LiteLLM-routed (text) requests carry a call id, so rows always have `modality=text`. operationId: getUsageRequest parameters: - name: request_id in: path required: true schema: type: string description: The `x-litellm-call-id` header value from the original response. responses: '200': description: Successful response content: application/json: schema: type: object required: - object - currency - request_id - model - canonical_model - status - prompt_tokens - completion_tokens - total_tokens - total_spend - modality - billing_unit - started_at - ended_at - generated_at properties: object: type: string enum: - usage.request currency: type: string enum: - USD request_id: type: string model: type: string description: >- Raw ledger id — the served routing target; not stable across packaging changes. canonical_model: type: string description: >- Stable catalog id; use this to group and reconcile per-model. status: type: string description: | `success`, `failure`, or `unknown`. Failed requests typically report zero tokens and $0 spend. prompt_tokens: type: integer completion_tokens: type: integer total_tokens: type: integer total_spend: type: number modality: type: string enum: - text units_consumed: type: number nullable: true billing_unit: type: string started_at: type: string format: date-time ended_at: type: string format: date-time generated_at: type: string format: date-time '401': $ref: '#/components/responses/Unauthorized' '404': description: | No request with this id is visible to this key — unknown id, a request made with a different key, or a row not yet written. content: application/json: schema: $ref: '#/components/schemas/Error' '503': description: Authentication service temporarily unavailable. content: application/json: schema: $ref: '#/components/schemas/Error' '504': description: Usage query timed out. Retry shortly. content: application/json: schema: $ref: '#/components/schemas/Error' components: responses: Unauthorized: description: Missing or invalid API key. content: application/json: schema: $ref: '#/components/schemas/Error' schemas: Error: type: object required: - error properties: error: type: object required: - message - type properties: message: type: string type: type: string example: invalid_request_error code: type: string description: | Stable machine-readable code for the failure. Present on most error paths; OMITTED on FastAPI default-validation bodies. Do not assume always present. doc_url: type: string format: uri description: | FlexAI extension. Link to the relevant dashboard / docs page for the failure mode (e.g. invalid key → /dashboard/keys). Present on auth and quota errors. securitySchemes: bearerAuth: type: http scheme: bearer bearerFormat: sk-xxxx description: | Virtual API key. Create one from the [FlexAI dashboard](https://tokens.flex.ai/dashboard/keys). Pass as `Authorization: Bearer sk-xxxx`. ````