A test run is a single execution of a benchmark test using a defined model configuration.
Each run represents how a particular large language model (LLM) — such as GPT-4, Claude-3, or Gemini — performed on a given task at a specific time, with specific settings.
A test run includes:
Together, test runs make it possible to compare models, providers, and configurations across benchmarks in a transparent and reproducible way.
newspaper-page document-understanding printed latin 20 en prose, columns
| Provider | mistral |
| Model | mistral-medium-3.5 |
| Temperature | 0.0 |
| Dataclass | MagazinePage |
| Normalized Score | 24.20 % |
| Test time | 58.04 seconds |
Extract all advertisements and return their bounding boxes.
The original size of the page is {width} x {height} pixels.
{"advertisements":[{"box":[100.0,50.0,2379.0,300.0]},{"box":[100.0,300.0,1180.0,500.0]},{"box":[1200.0,300.0,2379.0,500.0]},{"box":[100.0,500.0,2379.0,700.0]},{"box":[100.0,700.0,1180.0,900.0]},{"box":[1200.0,700.0,2379.0,900.0]},{"box":[100.0,900.0,1180.0,1100.0]},{"box":[1200.0,900.0,2379.0,1100.0]},{"box":[100.0,1100.0,1180.0,1300.0]},{"box":[1200.0,1100.0,2379.0,1300.0]}]}| Fuzzy Score | F1 micro / macro | Micro precision/recall | Tue/False Positives | |||||
| n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a |
| Micro Precision | Micro Recall | Instances | TP | FP | FN | |||
| Pricing Date: n/a, n/a. | Tokens: 88.4K IT + 2.9K OT = 91.3K TT | Cost: 0.133$ + 0.022$ = 0.155$ |
newspaper-page document-understanding printed latin 20 en prose, columns
| Provider | mistral |
| Model | ministral-14b-2512 |
| Temperature | 0.0 |
| Dataclass | MagazinePage |
| Normalized Score | 36.40 % |
| Test time | 3.80 seconds |
Extract all advertisements and return their bounding boxes.
The original size of the page is {width} x {height} pixels.
{"advertisements":[{"box":[220.0,250.0,2259.0,3300.0]}]}| Fuzzy Score | F1 micro / macro | Micro precision/recall | Tue/False Positives | |||||
| n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a |
| Micro Precision | Micro Recall | Instances | TP | FP | FN | |||
| Pricing Date: n/a, n/a. | Tokens: 112.9K IT + 4.8K OT = 117.8K TT | Cost: 0.023$ + 0.001$ = 0.024$ |
newspaper-page document-understanding printed latin 20 en prose, columns
| Provider | mistral |
| Model | mistral-large-2512 |
| Temperature | 0.0 |
| Dataclass | MagazinePage |
| Normalized Score | 18.10 % |
| Test time | 5.38 seconds |
Extract all advertisements and return their bounding boxes.
The original size of the page is {width} x {height} pixels.
{"advertisements":[{"box":[2000.0,120.0,2479.0,600.0]},{"box":[2000.0,650.0,2479.0,1200.0]},{"box":[2000.0,1800.0,2479.0,2400.0]}]}| Fuzzy Score | F1 micro / macro | Micro precision/recall | Tue/False Positives | |||||
| n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a |
| Micro Precision | Micro Recall | Instances | TP | FP | FN | |||
| Pricing Date: n/a, n/a. | Tokens: 78.6K IT + 3.3K OT = 81.8K TT | Cost: 0.039$ + 0.005$ = 0.044$ |
newspaper-page document-understanding printed latin 20 en prose, columns
| Provider | mistral |
| Model | ministral-8b-2512 |
| Temperature | 0.0 |
| Dataclass | MagazinePage |
| Normalized Score | 23.50 % |
| Test time | 1.87 seconds |
Extract all advertisements and return their bounding boxes.
The original size of the page is {width} x {height} pixels.
{"advertisements":[{"box":[200.0,100.0,2279.0,3408.0]}]}| Fuzzy Score | F1 micro / macro | Micro precision/recall | Tue/False Positives | |||||
| n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a |
| Micro Precision | Micro Recall | Instances | TP | FP | FN | |||
| Pricing Date: n/a, n/a. | Tokens: 112.9K IT + 9.8K OT = 122.7K TT | Cost: 0.017$ + 0.001$ = 0.018$ |
newspaper-page document-understanding printed latin 20 en prose, columns
| Provider | anthropic |
| Model | claude-opus-4-8 |
| Temperature | 0.0 |
| Dataclass | MagazinePage |
| Normalized Score | 81.60 % |
| Test time | 4.59 seconds |
Extract all advertisements and return their bounding boxes.
The original size of the page is {width} x {height} pixels.
{"advertisements":[{"box":[343.0,343.0,2287.0,1716.0]},{"box":[347.0,1761.0,1310.0,2110.0]},{"box":[1525.0,1761.0,2287.0,2110.0]},{"box":[347.0,2150.0,1310.0,2580.0]},{"box":[1360.0,2150.0,2287.0,2580.0]},{"box":[330.0,2620.0,2305.0,3225.0]}]}| Fuzzy Score | F1 micro / macro | Micro precision/recall | Tue/False Positives | |||||
| n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a |
| Micro Precision | Micro Recall | Instances | TP | FP | FN | |||
| Pricing Date: n/a, n/a. | Tokens: 213.6K IT + 4.2K OT = 217.8K TT | Cost: 1.068$ + 0.105$ = 1.173$ |
newspaper-page document-understanding printed latin 20 en prose, columns
| Provider | genai |
| Model | gemini-3.5-flash |
| Temperature | 0.0 |
| Dataclass | MagazinePage |
| Normalized Score | 68.30 % |
| Test time | 4.85 seconds |
Extract all advertisements and return their bounding boxes.
The original size of the page is {width} x {height} pixels.
{"advertisements":[{"box":[245.0,266.0,2315.0,3129.0]}]}| Fuzzy Score | F1 micro / macro | Micro precision/recall | Tue/False Positives | |||||
| n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a |
| Micro Precision | Micro Recall | Instances | TP | FP | FN | |||
| Pricing Date: n/a, n/a. | Tokens: 52.3K IT + 5.4K OT = 57.7K TT | Cost: 0.078$ + 0.048$ = 0.127$ |
newspaper-page document-understanding printed latin 20 en prose, columns
| Provider | openai |
| Model | gpt-5.5-2026-04-23 |
| Temperature | 1.0 |
| Dataclass | MagazinePage |
| Normalized Score | 95.60 % |
| Test time | 15.84 seconds |
Extract all advertisements and return their bounding boxes.
The original size of the page is {width} x {height} pixels.
{"advertisements":[{"box":[60.0,47.0,2431.0,3456.0]}]}| Fuzzy Score | F1 micro / macro | Micro precision/recall | Tue/False Positives | |||||
| n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a |
| Micro Precision | Micro Recall | Instances | TP | FP | FN | |||
| Pricing Date: n/a, n/a. | Tokens: 158.1K IT + 52.5K OT = 210.7K TT | Cost: 0.791$ + 1.575$ = 2.366$ |
newspaper-page document-understanding printed latin 20 en prose, columns
| Provider | openrouter |
| Model | qwen/qwen3.5-35b-a3b-20260224 |
| Temperature | 0.0 |
| Dataclass | MagazinePage |
| Normalized Score | 2.20 % |
| Test time | 5.42 seconds |
Extract all advertisements and return their bounding boxes.
The original size of the page is {width} x {height} pixels.
{"advertisements":[{"box":[146.0,106.0,929.0,926.0]}]}| Fuzzy Score | F1 micro / macro | Micro precision/recall | Tue/False Positives | |||||
| n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a |
| Micro Precision | Micro Recall | Instances | TP | FP | FN | |||
| Pricing Date: n/a, n/a. | Tokens: 132.7K IT + 26.6K OT = 159.2K TT | Cost: 0.022$ + 0.035$ = 0.056$ |
newspaper-page document-understanding printed latin 20 en prose, columns
| Provider | openrouter |
| Model | qwen/qwen3.5-9b-20260310 |
| Temperature | 0.0 |
| Dataclass | MagazinePage |
| Normalized Score | 5.40 % |
| Test time | 1498.97 seconds |
Extract all advertisements and return their bounding boxes.
The original size of the page is {width} x {height} pixels.
| Fuzzy Score | F1 micro / macro | Micro precision/recall | Tue/False Positives | |||||
| n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a |
| Micro Precision | Micro Recall | Instances | TP | FP | FN | |||
| Pricing Date: n/a, n/a. | Tokens: 127.4K IT + 507.3K OT = 634.8K TT | Cost: 0.013$ + 0.076$ = 0.089$ |
newspaper-page document-understanding printed latin 20 en prose, columns
| Provider | openrouter |
| Model | google/gemma-4-26b-a4b-it-20260403 |
| Temperature | 0.0 |
| Dataclass | MagazinePage |
| Normalized Score | 0.00 % |
| Test time | 1.12 seconds |
Extract all advertisements and return their bounding boxes.
The original size of the page is {width} x {height} pixels.
{"advertisements":[{"box":[135.0,143.0,885.0,911.0]}]}| Fuzzy Score | F1 micro / macro | Micro precision/recall | Tue/False Positives | |||||
| n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a |
| Micro Precision | Micro Recall | Instances | TP | FP | FN | |||
| Pricing Date: n/a, n/a. | Tokens: 15.8K IT + 3.9K OT = 19.7K TT | Cost: 0.001$ + 0.001$ = 0.003$ |