Sync MDPBench leaderboard from GitHub
Browse files- leaderboard.json +60 -30
- model_metadata.js +1 -0
leaderboard.json
CHANGED
|
@@ -2,7 +2,7 @@
|
|
| 2 |
"benchmark": {
|
| 3 |
"name": "MDPBench",
|
| 4 |
"version": "Initial official results",
|
| 5 |
-
"updated": "2026-08-
|
| 6 |
"metric": "Page-level mean of available text, formula, and table task scores (×100)",
|
| 7 |
"private_policy": "Private scores are run and verified by the MDPBench maintainers."
|
| 8 |
},
|
|
@@ -37,6 +37,36 @@
|
|
| 37 |
},
|
| 38 |
"source_name": "Gemini-3-pro-preview"
|
| 39 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 40 |
{
|
| 41 |
"type": "General VLM",
|
| 42 |
"name": "kimi-k2.5",
|
|
@@ -362,6 +392,35 @@
|
|
| 362 |
"ZH-T": 76.3
|
| 363 |
}
|
| 364 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 365 |
{
|
| 366 |
"type": "Specialized VLM",
|
| 367 |
"name": "PaddleOCR-VL-1.5",
|
|
@@ -479,35 +538,6 @@
|
|
| 479 |
"ZH-T": 76.2
|
| 480 |
}
|
| 481 |
},
|
| 482 |
-
{
|
| 483 |
-
"type": "Specialized VLM",
|
| 484 |
-
"name": "PaddleOCR-VL-1.6",
|
| 485 |
-
"overall": 75.0,
|
| 486 |
-
"digital": 82.8,
|
| 487 |
-
"photo": 72.6,
|
| 488 |
-
"latin": 78.0,
|
| 489 |
-
"non_latin": 71.6,
|
| 490 |
-
"private": 77.0,
|
| 491 |
-
"languages": {
|
| 492 |
-
"DE": 84.1,
|
| 493 |
-
"EN": 79.7,
|
| 494 |
-
"ES": 69.2,
|
| 495 |
-
"FR": 74.8,
|
| 496 |
-
"ID": 81.6,
|
| 497 |
-
"IT": 82.0,
|
| 498 |
-
"NL": 74.7,
|
| 499 |
-
"PT": 76.4,
|
| 500 |
-
"VI": 79.3,
|
| 501 |
-
"AR": 69.4,
|
| 502 |
-
"HI": 65.6,
|
| 503 |
-
"JP": 68.7,
|
| 504 |
-
"KO": 82.5,
|
| 505 |
-
"RU": 70.7,
|
| 506 |
-
"TH": 62.3,
|
| 507 |
-
"ZH": 78.0,
|
| 508 |
-
"ZH-T": 75.7
|
| 509 |
-
}
|
| 510 |
-
},
|
| 511 |
{
|
| 512 |
"type": "Specialized VLM",
|
| 513 |
"name": "MinerU-2.5-Pro-1.2B",
|
|
|
|
| 2 |
"benchmark": {
|
| 3 |
"name": "MDPBench",
|
| 4 |
"version": "Initial official results",
|
| 5 |
+
"updated": "2026-08-15",
|
| 6 |
"metric": "Page-level mean of available text, formula, and table task scores (×100)",
|
| 7 |
"private_policy": "Private scores are run and verified by the MDPBench maintainers."
|
| 8 |
},
|
|
|
|
| 37 |
},
|
| 38 |
"source_name": "Gemini-3-pro-preview"
|
| 39 |
},
|
| 40 |
+
{
|
| 41 |
+
"type": "General VLM",
|
| 42 |
+
"name": "Kimi-K3",
|
| 43 |
+
"overall": 83.6,
|
| 44 |
+
"digital": 90.8,
|
| 45 |
+
"photo": 81.2,
|
| 46 |
+
"latin": 86.2,
|
| 47 |
+
"non_latin": 80.7,
|
| 48 |
+
"private": 85.6,
|
| 49 |
+
"languages": {
|
| 50 |
+
"DE": 89.1,
|
| 51 |
+
"EN": 87.2,
|
| 52 |
+
"ES": 80.2,
|
| 53 |
+
"FR": 80.0,
|
| 54 |
+
"ID": 86.9,
|
| 55 |
+
"IT": 92.7,
|
| 56 |
+
"NL": 86.0,
|
| 57 |
+
"PT": 88.9,
|
| 58 |
+
"VI": 84.8,
|
| 59 |
+
"AR": 77.4,
|
| 60 |
+
"HI": 77.5,
|
| 61 |
+
"JP": 74.9,
|
| 62 |
+
"KO": 89.9,
|
| 63 |
+
"RU": 82.4,
|
| 64 |
+
"TH": 72.1,
|
| 65 |
+
"ZH": 89.5,
|
| 66 |
+
"ZH-T": 81.9
|
| 67 |
+
},
|
| 68 |
+
"source_name": "Kimi K3"
|
| 69 |
+
},
|
| 70 |
{
|
| 71 |
"type": "General VLM",
|
| 72 |
"name": "kimi-k2.5",
|
|
|
|
| 392 |
"ZH-T": 76.3
|
| 393 |
}
|
| 394 |
},
|
| 395 |
+
{
|
| 396 |
+
"type": "Specialized VLM",
|
| 397 |
+
"name": "PaddleOCR-VL-1.6",
|
| 398 |
+
"overall": 78.9,
|
| 399 |
+
"digital": 87.1,
|
| 400 |
+
"photo": 76.3,
|
| 401 |
+
"latin": 81.4,
|
| 402 |
+
"non_latin": 76.1,
|
| 403 |
+
"private": 83.0,
|
| 404 |
+
"languages": {
|
| 405 |
+
"DE": 85.1,
|
| 406 |
+
"EN": 83.5,
|
| 407 |
+
"ES": 75.1,
|
| 408 |
+
"FR": 78.6,
|
| 409 |
+
"ID": 82.7,
|
| 410 |
+
"IT": 85.5,
|
| 411 |
+
"NL": 80.5,
|
| 412 |
+
"PT": 80.9,
|
| 413 |
+
"VI": 80.9,
|
| 414 |
+
"AR": 76.2,
|
| 415 |
+
"HI": 70.7,
|
| 416 |
+
"JP": 70.3,
|
| 417 |
+
"KO": 86.3,
|
| 418 |
+
"RU": 75.1,
|
| 419 |
+
"TH": 67.3,
|
| 420 |
+
"ZH": 85.9,
|
| 421 |
+
"ZH-T": 76.7
|
| 422 |
+
}
|
| 423 |
+
},
|
| 424 |
{
|
| 425 |
"type": "Specialized VLM",
|
| 426 |
"name": "PaddleOCR-VL-1.5",
|
|
|
|
| 538 |
"ZH-T": 76.2
|
| 539 |
}
|
| 540 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 541 |
{
|
| 542 |
"type": "Specialized VLM",
|
| 543 |
"name": "MinerU-2.5-Pro-1.2B",
|
model_metadata.js
CHANGED
|
@@ -13,6 +13,7 @@ const MODEL_METADATA = {
|
|
| 13 |
{ label: "Hugging Face", url: "https://huggingface.co/moonshotai/Kimi-K2.5" }
|
| 14 |
]
|
| 15 |
},
|
|
|
|
| 16 |
"doubao-seed-2-0-pro-260215": { label: "Doubao-Seed-2.0-Pro (260215)" },
|
| 17 |
"claude-sonnet-4-6": { label: "Claude Sonnet 4.6" },
|
| 18 |
"gpt-5.2-2025-12-11": { label: "GPT-5.2 (2025-12-11)" },
|
|
|
|
| 13 |
{ label: "Hugging Face", url: "https://huggingface.co/moonshotai/Kimi-K2.5" }
|
| 14 |
]
|
| 15 |
},
|
| 16 |
+
"Kimi-K3": { label: "Kimi K3" },
|
| 17 |
"doubao-seed-2-0-pro-260215": { label: "Doubao-Seed-2.0-Pro (260215)" },
|
| 18 |
"claude-sonnet-4-6": { label: "Claude Sonnet 4.6" },
|
| 19 |
"gpt-5.2-2025-12-11": { label: "GPT-5.2 (2025-12-11)" },
|