Skip to content

Commit fa76234

Browse files
ericyangpanericyangpan
committed
feat(models): refresh intelligence catalog
1 parent 00c8f28 commit fa76234

9 files changed

Lines changed: 447 additions & 25 deletions

File tree

‎data/artificial-analysis-index.json‎

Lines changed: 34 additions & 16 deletions
Original file line numberDiff line numberDiff line change
@@ -4,14 +4,14 @@
44
"sourceUrl": "https://artificialanalysis.ai/leaderboards/models",
55
"methodologyUrl": "https://artificialanalysis.ai/methodology/intelligence-benchmarking",
66
"indexVersion": "4.1.1",
7-
"observedAt": "2026-08-19",
7+
"observedAt": "2026-08-28",
88
"legacyMissingModelIds": ["cursor-composer-2", "cursor-composer-2-5"],
99
"entries": [
1010
{
1111
"modelId": "claude-fable-5",
1212
"score": 62,
1313
"estimated": false,
14-
"configuration": "Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)"
14+
"configuration": "Claude Fable 5 (with fallback)"
1515
},
1616
{
1717
"modelId": "claude-haiku-3",
@@ -77,7 +77,7 @@
7777
"modelId": "claude-opus-5",
7878
"score": 63,
7979
"estimated": false,
80-
"configuration": "Claude Opus 5 (Adaptive Reasoning, Max Effort)"
80+
"configuration": "Claude Opus 5 (max)"
8181
},
8282
{
8383
"modelId": "claude-sonnet-3",
@@ -125,7 +125,7 @@
125125
"modelId": "claude-sonnet-5",
126126
"score": 55,
127127
"estimated": false,
128-
"configuration": "Claude Sonnet 5 (Adaptive Reasoning, Max Effort)"
128+
"configuration": "Claude Sonnet 5 (max)"
129129
},
130130
{
131131
"modelId": "deepseek-3-2",
@@ -173,7 +173,13 @@
173173
"modelId": "deepseek-v4-flash",
174174
"score": 52,
175175
"estimated": false,
176-
"configuration": "DeepSeek V4 Flash 0731 (Reasoning, Max Effort)"
176+
"configuration": "DeepSeek V4 Flash 0731 (max)"
177+
},
178+
{
179+
"modelId": "deepseek-v4-flash-vision",
180+
"score": 51,
181+
"estimated": false,
182+
"configuration": "DeepSeek V4 Flash Vision (max)"
177183
},
178184
{
179185
"modelId": "deepseek-v4-flash-preview",
@@ -185,13 +191,13 @@
185191
"modelId": "deepseek-v4-pro",
186192
"score": 53,
187193
"estimated": false,
188-
"configuration": "DeepSeek V4 Pro 0813 (Reasoning, Max Effort)"
194+
"configuration": "DeepSeek V4 Pro 0813 (max)"
189195
},
190196
{
191197
"modelId": "deepseek-v4-pro-preview",
192198
"score": 45,
193199
"estimated": false,
194-
"configuration": "DeepSeek V4 Pro (Reasoning, Max Effort)"
200+
"configuration": "DeepSeek V4 Pro (max)"
195201
},
196202
{
197203
"modelId": "devstral-2",
@@ -251,7 +257,7 @@
251257
"modelId": "gemini-3-6-flash",
252258
"score": 52,
253259
"estimated": false,
254-
"configuration": "Gemini 3.6 Flash (high)"
260+
"configuration": "Gemini 3.6 Flash"
255261
},
256262
{
257263
"modelId": "gemini-3-7-flash",
@@ -275,13 +281,13 @@
275281
"modelId": "gemma-4-26b-a4b",
276282
"score": 26,
277283
"estimated": false,
278-
"configuration": "Gemma 4 26B A4B (Reasoning)"
284+
"configuration": "Gemma 4 26B A4B"
279285
},
280286
{
281287
"modelId": "gemma-4-31b",
282288
"score": 30,
283289
"estimated": false,
284-
"configuration": "Gemma 4 31B (Reasoning)"
290+
"configuration": "Gemma 4 31B"
285291
},
286292
{
287293
"modelId": "glm-4-5",
@@ -361,6 +367,12 @@
361367
"estimated": false,
362368
"configuration": "GLM-5.3 (max)"
363369
},
370+
{
371+
"modelId": "glm-5-3-flash",
372+
"score": 57,
373+
"estimated": false,
374+
"configuration": "GLM-5.3-Flash"
375+
},
364376
{
365377
"modelId": "gpt-4-1",
366378
"score": 20,
@@ -659,7 +671,7 @@
659671
"modelId": "mistral-small-4",
660672
"score": 20,
661673
"estimated": false,
662-
"configuration": "Mistral Small 4 (Reasoning)"
674+
"configuration": "Mistral Small 4"
663675
},
664676
{
665677
"modelId": "muse-spark-1-1",
@@ -695,7 +707,7 @@
695707
"modelId": "qwen3-5-122b-a10b",
696708
"score": 33,
697709
"estimated": false,
698-
"configuration": "Qwen3.5 122B A10B (Reasoning)"
710+
"configuration": "Qwen3.5 122B A10B"
699711
},
700712
{
701713
"modelId": "qwen3-5-35b-a3b",
@@ -707,7 +719,7 @@
707719
"modelId": "qwen3-5-397b-a17b",
708720
"score": 34,
709721
"estimated": false,
710-
"configuration": "Qwen3.5 397B A17B (Reasoning)"
722+
"configuration": "Qwen3.5 397B A17B"
711723
},
712724
{
713725
"modelId": "qwen3-coder-30b-a3b",
@@ -731,13 +743,13 @@
731743
"modelId": "qwen3-6-27b",
732744
"score": 38,
733745
"estimated": false,
734-
"configuration": "Qwen3.6 27B (Reasoning)"
746+
"configuration": "Qwen3.6 27B"
735747
},
736748
{
737749
"modelId": "qwen3-6-35b-a3b",
738750
"score": 32,
739751
"estimated": false,
740-
"configuration": "Qwen3.6 35B A3B (Reasoning)"
752+
"configuration": "Qwen3.6 35B A3B"
741753
},
742754
{
743755
"modelId": "qwen3-6-max-preview",
@@ -767,14 +779,20 @@
767779
"modelId": "qwen3-8-27b",
768780
"score": 52,
769781
"estimated": false,
770-
"configuration": "Qwen3.8 27B"
782+
"configuration": "Qwen3.8 27B (xhigh)"
771783
},
772784
{
773785
"modelId": "qwen3-8-max",
774786
"score": 58,
775787
"estimated": false,
776788
"configuration": "Qwen3.8 Max"
777789
},
790+
{
791+
"modelId": "qwen3-8-flash-next",
792+
"score": 56,
793+
"estimated": false,
794+
"configuration": "Qwen3.8-Flash-Next"
795+
},
778796
{
779797
"modelId": "qwen3-coder-next",
780798
"score": 21,

‎manifests/models/deepseek-v4-flash-vision.json‎

Lines changed: 110 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,110 @@
1+
{
2+
"$schema": "../$schemas/model.schema.json",
3+
"id": "deepseek-v4-flash-vision",
4+
"name": "DeepSeek-V4-Flash-Vision-Exp",
5+
"description": "DeepSeek's experimental multimodal V4 Flash model for text and image understanding, with thinking enabled by default, a 1M-token context window, and up to 384K output tokens.",
6+
"translations": {
7+
"de": {
8+
"description": "DeepSeeks experimentelles multimodales V4-Flash-Modell für Text- und Bildverständnis, mit standardmäßig aktivem Reasoning, 1 Mio. Token Kontext und bis zu 384K Ausgabetokens."
9+
},
10+
"es": {
11+
"description": "Modelo V4 Flash multimodal experimental de DeepSeek para texto e imágenes, con razonamiento activo por defecto, contexto de 1 M y hasta 384K tokens de salida."
12+
},
13+
"fr": {
14+
"description": "Modèle V4 Flash multimodal expérimental de DeepSeek pour le texte et l’image, avec raisonnement activé par défaut, contexte de 1 M et jusqu’à 384K jetons de sortie."
15+
},
16+
"id": {
17+
"description": "Model V4 Flash multimodal eksperimental DeepSeek untuk memahami teks dan gambar, dengan penalaran aktif secara default, konteks 1 juta token, dan output hingga 384K token."
18+
},
19+
"ja": {
20+
"description": "テキストと画像を理解し、推論が既定で有効、100万トークン文脈と最大384K出力を備えたDeepSeekの実験的マルチモーダルV4 Flashモデル。"
21+
},
22+
"ko": {
23+
"description": "텍스트와 이미지를 이해하고 추론이 기본 활성화되며 100만 토큰 컨텍스트와 최대 384K 출력을 지원하는 DeepSeek의 실험적 멀티모달 V4 Flash 모델입니다."
24+
},
25+
"pt": {
26+
"description": "Modelo V4 Flash multimodal experimental da DeepSeek para texto e imagens, com raciocínio ativado por padrão, contexto de 1 milhão e até 384K tokens de saída."
27+
},
28+
"ru": {
29+
"description": "Экспериментальная мультимодальная V4 Flash от DeepSeek для текста и изображений: рассуждение по умолчанию, контекст 1 млн и вывод до 384K токенов."
30+
},
31+
"tr": {
32+
"description": "DeepSeek'in metin ve görüntü anlayışı için deneysel multimodal V4 Flash modeli; varsayılan açık akıl yürütme, 1 milyon token bağlam ve 384K'ya kadar çıktı sunar."
33+
},
34+
"zh-Hans": {
35+
"description": "DeepSeek 面向文本与图像理解的实验性多模态 V4 Flash 模型，默认启用推理，支持 100 万 token 上下文和最高 384K 输出。"
36+
},
37+
"zh-Hant": {
38+
"description": "DeepSeek 面向文字與圖像理解的實驗性多模態 V4 Flash 模型，預設啟用推理，支援 100 萬 token 上下文和最高 384K 輸出。"
39+
}
40+
},
41+
"verified": true,
42+
"sources": [
43+
{
44+
"url": "https://api-docs.deepseek.com/updates/",
45+
"title": "DeepSeek API change log",
46+
"fields": [
47+
"name",
48+
"description",
49+
"websiteUrl",
50+
"docsUrl",
51+
"releaseDate",
52+
"lifecycle",
53+
"inputModalities",
54+
"outputModalities",
55+
"capabilities"
56+
]
57+
},
58+
{
59+
"url": "https://api-docs.deepseek.com/quick_start/pricing",
60+
"title": "DeepSeek models and pricing",
61+
"fields": ["contextWindow", "maxOutput", "tokenPricing", "capabilities"]
62+
},
63+
{
64+
"url": "https://api-docs.deepseek.com/guides/vision",
65+
"title": "DeepSeek Vision guide",
66+
"fields": ["inputModalities", "outputModalities"]
67+
},
68+
{
69+
"url": "https://artificialanalysis.ai/models/deepseek-v4-flash-vision",
70+
"title": "Artificial Analysis DeepSeek V4 Flash Vision model page",
71+
"fields": ["platformUrls.artificialAnalysis"]
72+
}
73+
],
74+
"lastVerifiedAt": "2026-08-28",
75+
"verifiedBy": "codex-agent",
76+
"confidence": "high",
77+
"websiteUrl": "https://www.deepseek.com",
78+
"docsUrl": "https://api-docs.deepseek.com/guides/vision",
79+
"vendor": "DeepSeek",
80+
"size": null,
81+
"activeParameters": null,
82+
"contextWindow": 1000000,
83+
"maxOutput": 393216,
84+
"tokenPricing": {
85+
"status": "unavailable",
86+
"reason": "unsupported-pricing-structure",
87+
"primaryOffer": null,
88+
"offers": []
89+
},
90+
"releaseDate": "2026-08-21",
91+
"lifecycle": "latest",
92+
"knowledgeCutoff": null,
93+
"inputModalities": ["text", "image"],
94+
"outputModalities": ["text"],
95+
"capabilities": ["function-calling", "tool-choice", "structured-outputs", "reasoning"],
96+
"benchmarks": {
97+
"sweBench": null,
98+
"terminalBench": null,
99+
"mmmu": null,
100+
"mmmuPro": null,
101+
"webDevArena": null,
102+
"sciCode": null,
103+
"liveCodeBench": null
104+
},
105+
"platformUrls": {
106+
"huggingface": null,
107+
"artificialAnalysis": "https://artificialanalysis.ai/models/deepseek-v4-flash-vision",
108+
"openrouter": null
109+
}
110+
}

0 commit comments

Comments
 (0)