← सभी बेंचमार्क

टेक्स्ट एरिना

Wikiprompt से, प्रॉम्प्ट का मुक्त विश्वकोश

सामुदायिक Elo रैंकिंग, LMArena की लड़ाइयों से दैनिक रूप से एकत्रित। पहले सामान्य एरिना, फिर श्रेणी के अनुसार विवरण।

अंतिम अपडेट: 2026-11-09 · 217 मॉडल

स्थानमॉडलसंगठनElo स्कोर
5Claude Opus 5 (Max)anthropic1,506
7Claude Opus 5 (High)anthropic1,502
8Claude Opus 4.6 (High)anthropic1,502
9Gemini 3.8 Flash (High)google1,500
10Claude Opus 4.6anthropic1,497
12Claude Fable 5 (High)anthropic1,491
13Muse Spark 1.3 (Max)meta1,491
14Claude Opus 4.7 (High)anthropic1,490
15Gemini 3.7 Flash (High)google1,489
16Qwen3.8-Maxalibaba1,485
18Claude Opus 4.7anthropic1,483
19Gemini 3.5 Flash (High)google1,483
20Gemini 3.1 Pro Previewgoogle1,480
21Gemini 3 Progoogle1,480
22Gemini 3.6 Flash (High)google1,479
23Muse Spark 1.1meta1,479
24Gemini 3.5 Flash (Medium)google1,478
25GLM-5.3 (Max)zai1,476
27Kimi K3 (Max)moonshot1,475
29GLM-5.3-Flashzai1,471
30GPT-5.5 (High)openai1,471
31Qwen3.5-Max-Previewalibaba1,471
32GLM-5.2 (Max)zai1,469
33GPT-5.4 (High)openai1,469
35ERNIE 5.1baidu1,468
36GPT-5.5openai1,466
37Gemini 3 Flashgoogle1,466
38MiMo-V2.5-Proxiaomi1,464
40GLM-5.1zai1,463
41Claude Opus 4.8 (High)anthropic1,462
42Gemini 2.5 Progoogle1,458
43Claude Sonnet 4.6anthropic1,458
44GPT-5.6 Sol (xHigh)openai1,457
45Kimi K2.6moonshot1,455
46Qwen3.7-Plusalibaba1,454
47Claude Opus 4.8anthropic1,454
48GPT-5.4openai1,452
49DeepSeek-V4-Prodeepseek1,451
50Grok 4.20 Beta (Reasoning, 0309)xai1,451
51Claude Opus 4.5 (20251101)anthropic1,451
52Grok 4.20 Multi-Agent Beta (0309)xai1,451
53MiMo-V2.6-Flashxiaomi1,451
54GPT-5.6 Terra (xHigh)openai1,449
56Dola Seed 2.0 Probytedance1,449
57Claude Opus 4.5 (High, 32K, 20251101)anthropic1,448
58Grok 4.5xai1,448
59GLM-5zai1,446
62DeepSeek-V4-Pro Preview (High)deepseek1,445
65Kimi K2.5 (Thinking)moonshot1,445
66Grok 4.20 Beta 1xai1,444
67ERNIE 5.0 (0110)baidu1,444
68Inklingthinky1,443
70GPT-5.1 (High)openai1,443
71Gemini 3 Flash (Minimal)google1,443
74Claude Sonnet 5 (High)anthropic1,442
75GLM-4.6zai1,440
77Qwen3-Max-Previewalibaba1,439
79GPT-5.2 Chat (20260210)openai1,439
80Claude Sonnet 4.5 (20250929)anthropic1,439
82Qwen3.5-397B-A17Balibaba1,438
83Grok 4.1 (Thinking)xai1,437
84MiMo-V2-Proxiaomi1,437
85Qwen3.8-27Balibaba1,436
86Qwen3.6-Plusalibaba1,436
87Grok 4.1xai1,436
89Claude Sonnet 4.5 (High, 32K, 20250929)anthropic1,435
90Gemini 3.5 Flash-Litegoogle1,434
92DeepSeek-V4-Flashdeepseek1,432
93GPT-5.6 Luna (xHigh)openai1,431
94GLM-4.5zai1,430
95Grok 4.6 (High)xai1,430
97Mistral Large 3mistral1,429
98ChatGPT-4o Latest (20250326)openai1,429
99MiniMax-M3minimax1,428
101MiMo-V2.5xiaomi1,428
102LongCat-Flash-Chat Exp (2602)meituan1,426
103Grok 3 Preview (02-24)xai1,426
106DeepSeek-V3.2deepseek1,425
107Mistral Medium 3.1 (2508)mistral1,425
108DeepSeek-V4-Flash Preview (High)deepseek1,424
112GPT-5.1openai1,422
115DeepSeek-V3.2 (Thinking)deepseek1,420
118GPT-5.5 Instantopenai1,419
121Claude Opus 4.1 (Thinking 16K, 20250805)anthropic1,419
122Qwen3-235B-A22B-Instruct (2507)alibaba1,419
123Claude Opus 4.1 (20250805)anthropic1,419
125Qwen3.5-122B-A10Balibaba1,417
127Qwen3-Next-80B-A3B-Instructalibaba1,417
128Gemini 2.5 Flashgoogle1,417
130GPT-5.2 (High)openai1,416
131Gemini 3.1 Flash-Lite Previewgoogle1,416
133amazon-nova-experimental-chat (11-10)amazon1,415
134Kimi K2 Thinking Turbomoonshot1,414
135Inkling-Smallthinky1,414
137GPT-5.2openai1,412
138GPT-5.4 mini (High)openai1,412
139Grok 4 (0709)xai1,411
140MiMo-V2-Flash (Non-Thinking)xiaomi1,411
142o3 (2025-04-16)openai1,410
143Qwen3.5-27Balibaba1,409

श्रेणी के अनुसार

Exclude ties

स्थानमॉडलसंगठनElo स्कोर
5Claude Opus 5 (Max)anthropic1,524
7Claude Opus 5 (High)anthropic1,518
8Claude Opus 4.6 (High)anthropic1,517
9Gemini 3.8 Flash (High)google1,516
10Claude Opus 4.6anthropic1,512
11Claude Fable 5 (High)anthropic1,504
14Claude Opus 4.7 (High)anthropic1,501
15Gemini 3.7 Flash (High)google1,500
16Qwen3.8-Maxalibaba1,497
18Claude Opus 4.7anthropic1,492
19Gemini 3.5 Flash (High)google1,491
20Gemini 3 Progoogle1,491
21Gemini 3.1 Pro Previewgoogle1,490
22Gemini 3.6 Flash (High)google1,488
23Muse Spark 1.1meta1,486

Non english

स्थानमॉडलसंगठनElo स्कोर
5Claude Opus 5 (Max)anthropic1,500
6Claude Opus 5 (High)anthropic1,495
7Gemini 3.8 Flash (High)google1,493
9Claude Opus 4.6 (High)anthropic1,487
11Claude Opus 4.6anthropic1,484
13Claude Fable 5 (High)anthropic1,482
14Claude Opus 4.7 (High)anthropic1,480
16Gemini 3.5 Flash (High)google1,477
17Gemini 3.1 Pro Previewgoogle1,476
19Qwen3.8-Maxalibaba1,475
20Gemini 3 Progoogle1,474
21Claude Opus 4.7anthropic1,474
23Gemini 3.5 Flash (Medium)google1,472
24Muse Spark 1.1meta1,472
26GPT-5.5 (High)openai1,467

Hard prompts

स्थानमॉडलसंगठनElo स्कोर
5Claude Opus 5 (Max)anthropic1,525
6Claude Opus 4.6 (High)anthropic1,525
7Claude Opus 5 (High)anthropic1,522
8Claude Opus 4.6anthropic1,522
11Gemini 3.8 Flash (High)google1,511
12Claude Fable 5 (High)anthropic1,508
13Claude Opus 4.7 (High)anthropic1,506
15Claude Opus 4.7anthropic1,500
16Qwen3.8-Maxalibaba1,499
17Gemini 3.7 Flash (High)google1,498
18Kimi K3 (Max)moonshot1,494
21GLM-5.3-Flashzai1,490
22Gemini 3.5 Flash (High)google1,489
23GPT-5.5 (High)openai1,488
25Muse Spark 1.1meta1,486

English

स्थानमॉडलसंगठनElo स्कोर
4Claude Opus 4.6 (High)anthropic1,512
8Claude Opus 4.6anthropic1,506
10Claude Opus 5 (High)anthropic1,502
11Gemini 3.8 Flash (High)google1,500
13Claude Opus 4.7 (High)anthropic1,496
16Claude Opus 4.7anthropic1,487
25Gemini 3.1 Pro Previewgoogle1,478
29MiMo-V2.5-Proxiaomi1,477
32GLM-5.1zai1,475
34GLM-5.2 (Max)zai1,475
37Claude Sonnet 4.6anthropic1,473
38GPT-5.5 (High)openai1,470
40Claude Opus 4.8 (High)anthropic1,469
41GPT-5.5openai1,469
43GPT-5.4 (High)openai1,467

Instruction following

स्थानमॉडलसंगठनElo स्कोर
4Claude Opus 4.6 (High)anthropic1,521
7Claude Opus 5 (High)anthropic1,515
8Claude Opus 4.6anthropic1,510
10Claude Opus 4.7 (High)anthropic1,498
12Gemini 3.8 Flash (High)google1,494
14Claude Opus 4.7anthropic1,489
23GPT-5.5 (High)openai1,478
25Claude Opus 4.8 (High)anthropic1,477
26Claude Sonnet 4.6anthropic1,476
27MiMo-V2.5-Proxiaomi1,476
28GPT-5.5openai1,475
30Claude Opus 4.5 (20251101)anthropic1,472
31GPT-5.4 (High)openai1,469
38Gemini 3.1 Pro Previewgoogle1,465
40Claude Opus 4.8anthropic1,463

Industry software and it services

स्थानमॉडलसंगठनElo स्कोर
5Claude Opus 4.6anthropic1,527
6Claude Opus 5 (High)anthropic1,526
8Claude Opus 4.6 (High)anthropic1,526
11Claude Opus 4.7 (High)anthropic1,516
13Gemini 3.8 Flash (High)google1,509
16Claude Opus 4.7anthropic1,507
26MiMo-V2.5-Proxiaomi1,495
30GPT-5.5 (High)openai1,492
32GPT-5.4 (High)openai1,492
33Claude Sonnet 4.6anthropic1,491
34Claude Opus 4.8 (High)anthropic1,488
39GPT-5.5openai1,486
41Gemini 3.1 Pro Previewgoogle1,486
44GLM-5.1zai1,484
45Claude Opus 4.5 (20251101)anthropic1,483

Longer query

स्थानमॉडलसंगठनElo स्कोर
5Claude Opus 4.6 (High)anthropic1,518
7Claude Opus 4.6anthropic1,515
8Claude Opus 5 (High)anthropic1,512
10Gemini 3.8 Flash (High)google1,509
11Claude Opus 4.7 (High)anthropic1,505
14Claude Opus 4.7anthropic1,499
21Claude Opus 4.8 (High)anthropic1,484
22GPT-5.5 (High)openai1,484
23Gemini 3.5 Flash (High)google1,484
24Gemini 3.1 Pro Previewgoogle1,482
25GPT-5.5openai1,482
26GLM-5.3-Flashzai1,482
28MiMo-V2.5-Proxiaomi1,481
30Claude Opus 4.8anthropic1,479
31Claude Opus 4.5 (20251101)anthropic1,479

Industry writing and literature and language

स्थानमॉडलसंगठनElo स्कोर
6Claude Opus 4.6 (High)anthropic1,502
12Claude Opus 4.6anthropic1,492
14Gemini 3.1 Pro Previewgoogle1,481
38Gemini 2.5 Progoogle1,454
59Gemini 3 Flash (Minimal)google1,436
85Qwen3.5-397B-A17Balibaba1,415
108Gemini 2.5 Flashgoogle1,405
118Mistral Large 3mistral1,397
121Mistral Medium 3.1 (2508)mistral1,396
133Qwen3-235B-A22B-Instruct (2507)alibaba1,390
139Claude Haiku 4.5 (20251001)anthropic1,386
149GPT-6 Luna (Max)openai1,380
156MiniMax-M2.7minimax1,376
220Claude 3.5 Sonnet (20241022)anthropic1,314
236GPT-4o (2024-05-13)openai1,298

Hard prompts english

स्थानमॉडलसंगठनElo स्कोर
4Claude Opus 4.6 (High)anthropic1,531
5Claude Opus 4.6anthropic1,528
18MiMo-V2.5-Proxiaomi1,494
30GPT-5.5openai1,483
38Gemini 3.1 Pro Previewgoogle1,480
79Gemini 2.5 Progoogle1,452
86Dola Seed 2.0 Probytedance1,450
88Qwen3.5-397B-A17Balibaba1,449
105MiniMax-M2.7minimax1,440
112Mistral Large 3mistral1,438
116Qwen3-235B-A22B-Instruct (2507)alibaba1,436
119Claude Haiku 4.5 (20251001)anthropic1,435
120Mistral Medium 3.1 (2508)mistral1,435
122Gemini 3 Flash (Minimal)google1,433
142GPT-5.2openai1,423

Coding

स्थानमॉडलसंगठनElo स्कोर
4Claude Opus 4.6anthropic1,534
8Claude Opus 4.6 (High)anthropic1,531
53Gemini 3.1 Pro Previewgoogle1,482
66Dola Seed 2.0 Probytedance1,472
82Qwen3.5-397B-A17Balibaba1,463
95MiniMax-M2.7minimax1,454
98Claude Haiku 4.5 (20251001)anthropic1,453
102Gemini 2.5 Progoogle1,452
104Mistral Large 3mistral1,449
114Qwen3-235B-A22B-Instruct (2507)alibaba1,445
127Gemini 3 Flash (Minimal)google1,436
133Mistral Medium 3.1 (2508)mistral1,434
138GPT-5.2openai1,432
139GPT-6 Luna (Max)openai1,432
149Gemini 2.5 Flashgoogle1,423

Industry entertainment and sports and media

स्थानमॉडलसंगठनElo स्कोर
16Gemini 3.1 Pro Previewgoogle1,465
38Gemini 2.5 Progoogle1,442
103Qwen3.5-397B-A17Balibaba1,398
113Gemini 2.5 Flashgoogle1,393
143Claude Haiku 4.5 (20251001)anthropic1,369
160GPT-6 Luna (Max)openai1,358
293Claude 3 Opus (20240229)anthropic1,220
307Llama 3 70B Instructmeta1,195

Creative writing

स्थानमॉडलसंगठनElo स्कोर
14Gemini 3.1 Pro Previewgoogle1,483
144Claude Haiku 4.5 (20251001)anthropic1,371
159GPT-6 Luna (Max)openai1,358
289Claude 3 Opus (20240229)anthropic1,235
304Llama 3 70B Instructmeta1,210

Multi turn

स्थानमॉडलसंगठनElo स्कोर
19Gemini 3.1 Pro Previewgoogle1,486
140Claude Haiku 4.5 (20251001)anthropic1,409
144Gemini 2.5 Flashgoogle1,408
273Claude 3 Opus (20240229)anthropic1,275
304Llama 3 70B Instructmeta1,223

Industry business and management and financial operations

स्थानमॉडलसंगठनElo स्कोर
35Gemini 3.1 Pro Previewgoogle1,459
59Gemini 2.5 Progoogle1,440
136Gemini 2.5 Flashgoogle1,404
142Claude Haiku 4.5 (20251001)anthropic1,399
298Claude 3 Opus (20240229)anthropic1,222

Industry life and physical and social science

स्थानमॉडलसंगठनElo स्कोर
22Gemini 3.1 Pro Previewgoogle1,496
165Claude Haiku 4.5 (20251001)anthropic1,406
295Claude 3 Opus (20240229)anthropic1,254
313Llama 3 70B Instructmeta1,226

Math

स्थानमॉडलसंगठनElo स्कोर
262Claude 3 Opus (20240229)anthropic1,273

Chinese

स्थानमॉडलसंगठनElo स्कोर
274Claude 3 Opus (20240229)anthropic1,248

Industry mathematical

स्थानमॉडलसंगठनElo स्कोर
276Claude 3 Opus (20240229)anthropic1,264

यह भी देखें

स्रोत और पद्धति

रैंकिंग सामुदायिक खुली मूल्यांकन से एकत्रित की जाती हैं। Elo स्कोर, विश्वास अंतराल और वोट LMArena से आते हैं (सामुदायिक मतदान वाले मॉडल युद्ध, खुले डेटासेट के रूप में प्रकाशित)। मूल्य और संदर्भ मेटाडेटा OpenRouter की सार्वजनिक API से आते हैं। Wikiprompt दैनिक स्नैपशॉट लेता है और इतिहास संरक्षित करता है; हम अपना मूल्यांकन नहीं चलाते।

  1. LMArena leaderboard dataset (Hugging Face)
  2. LMArena leaderboard