# /geekbench/ai_results

`POST /api/geekbench/ai_results`

Price: 1 credit

Fetch one uploaded Geekbench AI run by its id. Returns the single-precision, half-precision and quantized scores, the inference framework and backend the run used, and the per-workload breakdown — each of the ten workloads at all three precisions, with its score, its throughput figure and the accuracy it achieved. The host machine comes with it: operating system, model and model identifier, motherboard, processor name, topology, clocks and caches, and memory size.

## How to use it

Geekbench AI has one version and its own id space, so an id taken from a CPU or GPU run does not resolve here. One run measures one backend, so the same machine appears separately for its CPU, GPU and Neural Engine paths and the scores are not comparable across them — read `backend` before ranking anything. `accuracy` on a workload is the share of reference outputs matched at that precision, and it is the reason a quantized score far above the single-precision one is not simply better.

## Parameters

- `access-token` (string, required)

## Request body

- `timeout` (integer) — Max scrapping execution timeout (in seconds) (default: 300; min: 20; max: 1500)
- `id` (string, required) — Numeric id of the uploaded Geekbench AI run, as it appears in the browser.geekbench.com URL (examples: "560148", "559854"; minLength: 1)

## Response

### 200 — Successful Response

- `@type` (string) (default: "GeekbenchAiResult")
- `id` (string, required)
- `url` (string, required)
- `version` (string, required)
- `system_title` (string, nullable)
- `uploaded_at` (integer, nullable)
- `view_count` (integer, nullable)
- `user` (string, nullable)
- `benchmark_version` (string, nullable)
- `validation` (string, nullable)
- `operating_system` (string, nullable)
- `system_model` (string, nullable)
- `system_model_id` (string, nullable)
- `motherboard` (string, nullable)
- `cpu_name` (string, nullable)
- `cpu_identifier` (string, nullable)
- `cpu_topology` (string, nullable)
- `cpu_processor_count` (integer, nullable)
- `cpu_core_count` (integer, nullable)
- `cpu_thread_count` (integer, nullable)
- `cpu_base_frequency_mhz` (integer, nullable)
- `cpu_max_frequency_mhz` (integer, nullable)
- `memory_size_gb` (number, nullable)
- `memory_type` (string, nullable)
- `memory_channel_count` (integer, nullable)
- `scores` (object) (default: {})
  - `*` (integer)
- `information` (array) (default: [])
  - `@type` (string) (default: "GeekbenchInfoSection")
  - `name` (string, required)
  - `details` (object) (default: {})
    - `*` (string)
- `benchmarks` (array) (default: [])
  - `@type` (string) (default: "GeekbenchBenchmarkSection")
  - `name` (string, required)
  - `score` (integer, nullable)
  - `workloads` (array) (default: [])
    - `@type` (string) (default: "GeekbenchWorkload")
    - `name` (string, required)
    - `score` (integer, nullable)
    - `measurement` (string, nullable)
    - `accuracy` (string, nullable)
- `framework` (string, nullable)
- `backend` (string, nullable)
- `device` (string, nullable)
- `single_precision_score` (integer, nullable)
- `half_precision_score` (integer, nullable)
- `quantized_score` (integer, nullable)

## Errors

### 422 — Validation Error

The request body did not validate

What to do: Check the fields against this schema. A URN with the wrong prefix is the most common cause.

- `detail` (array)
  - `loc` (array, required)
  - `msg` (string, required)
  - `type` (string, required)
  - `input` (any)
  - `ctx` (object)

### 408

The request ran past its time limit

What to do: Raise `timeout` in the request body, up to the maximum this endpoint documents. Lowering `count` or turning off the `with_*` flags also helps, because less work finishes sooner.

### 412

No Geekbench AI run with that id, or the upload is withheld because it ran on unreleased hardware

What to do: Retrying will not help: either the entity does not exist, or the input points at a different one.

### 429

Too many requests: a rate limit or a usage window is exhausted

What to do: When the response carries an X-Retry-After header, wait that many seconds and retry: the same number is in the body as `detail.retry_after`, and the limit clears once that window passes. The message in the body names the limit that was hit.

### 500

Something broke on our side

What to do: Retrying will not help. If it keeps happening, send us the X-Request-ID from the response headers.

### 529

Rate limit reached, or the endpoint is overloaded

What to do: Wait at least 30 seconds, then retry.

## Response envelope

Success: Array of objects (may be empty if no results)

Error: Error may coexist with partial results if it occurs mid-execution. Check X-Error header and status code.

Every response carries these headers:

- `X-Error` — Error message text (present only on error)
- `X-Request-ID` — Unique request identifier
- `X-Execution-Time` — Execution time in seconds
- `X-Result-Count` — How many records the body carries. 0 means an empty result, which is a normal answer and not by itself an error. A non-zero count can come back together with X-Error when the failure happened partway through — read this header and X-Error independently.
- `X-Total-Available-Results` — How many records exist for this query, when the endpoint can say. On a `dry_run` request this is the answer and the body is empty. It saturates: the endpoint's documented maximum means 'at least that many', any smaller number is exact.
- `X-Warning` — Present when the request body carried keys this endpoint does not document. They were ignored, so any filter you meant to apply through them did not apply. Check the spelling against this schema and retry.
- `X-Retry-After` — Seconds to wait before retrying. Present only on 429.