Oneable ข่าวและความรู้ AI · ซอฟต์แวร์เฮาส์สำหรับธุรกิจไทย AI news & know-how · Software house for Thai businesses
AI Benchmarks

เทียบโมเดล AI ด้วย Benchmark จริงAI model benchmarks

คะแนนความฉลาด การเขียนโค้ด การทำงานแบบ agent การค้นเว็บ และงานออกแบบ รวมจาก Artificial Analysis, Design Arena และการทดสอบของ OpenRouter — อัปเดตทุกคืนIntelligence, coding, agentic, web-search and design scores from Artificial Analysis, Design Arena and OpenRouter evals — refreshed nightly.

ข้อมูลล่าสุดLast updated 1 Nov 2026

ดัชนีความฉลาด (Intelligence Index)Intelligence Index

คะแนนรวมจากชุดทดสอบความรู้ การใช้เหตุผล และคณิตศาสตร์ ยิ่งสูงยิ่งเก่ง · ราคาเป็น $ ต่อ 1 ล้าน token (input / output)Composite of knowledge, reasoning and maths evals — higher is better · price in $ per 1M tokens (input / output)

  1. 1 Claude Opus 5.5 (Max, Default Fallback) $4.4 / $22 58.4
  2. 2 GPT-6 Astra (Max) $10 / $50 · coding 76.9 · agentic 51 56.9
  3. 3 GPT-6.1 Sol (Max) $2 / $10 54.5
  4. 4 Claude Sonnet 5.5 (Max, Default Fallback) $2.2 / $11 54.2
  5. 5 Qwen3.8 Max · coding 68.9 · agentic 49.9 53.4
  6. 6 GPT-6.1 Terra (Max) $1.2 / $6 51
  7. 7 Gemini 4 Argon (High) $2.2 / $11 50.4
  8. 8 Claude Fable 5.1 (Xhigh, Default Fallback) $5 / $25 · coding 80.7 · agentic 57.2 49.7
  9. 9 Kimi K3.1 Pro $4 / $20 46
  10. 10 Muse Spark 1.3 (Max) $1.25 / $4.25 · coding 75.8 · agentic 55.5 36.4
  11. 11 Claude Haiku 5.5 (Max) $0.11 / $0.55 36.2
  12. 12 Grok 4.7 (High) $2.2 / $6.6 35
  13. 13 Qwen3.8 Max (0902) $2 / $6 · coding 76.2 · agentic 56 33.6
  14. 14 GLM-5.3 (Max) $0.98 / $3.08 · coding 74.8 · agentic 53.1 31.7
  15. 15 Gemini 3.8 Flash (High) $0.83 / $4.13 · coding 76.3 · agentic 40.2 30.5
  16. 16 MiMo-V2.6-Pro $0.52 / $1.04 29.9
  17. 17 MiniMax-M3.1-Flash $0.14 / $0.54 28.9
  18. 18 Step 5 Preview $1 / $2.7 28.5
  19. 19 Solar Pro 4 $0.15 / $0.6 · coding 52.7 28.2
  20. 20 GLM-5.3-Flash $0.15 / $0.5 · coding 71.5 · agentic 50.9 27.8
  21. 21 GPT-6 Luna (Max) $0.1 / $0.5 27.8
  22. 22 MiMo-V2.6-Flash $0.14 / $0.28 26.7
  23. 23 Mistral Large 4 (High) $0.77 / $2.31 26
  24. 24 DeepSeek V4 Pro 0813 (Max) $0.12 / $4 · coding 68.8 · agentic 41.3 23.1
  25. 25 Qwen3.8 27B (Xhigh) $0.04 / $2.3 · coding 68.1 · agentic 45.8 20.1

Source: Artificial Analysis (artificialanalysis.ai) via OpenRouter (openrouter.ai/rankings). · as of 30 Oct 2026

ดัชนีการเขียนโค้ด (Coding Index)Coding Index

คะแนนจากชุดทดสอบการเขียนและแก้โค้ดScores from code-writing and code-editing evals

  1. 1 Claude Fable 5.1 (Xhigh, Default Fallback) $5 / $25 per 1M 80.7
  2. 2 GPT-6 Astra (Max) $10 / $50 per 1M 76.9
  3. 3 Gemini 3.8 Flash (High) $0.83 / $4.13 per 1M 76.3
  4. 4 Qwen3.8 Max (0902) $2 / $6 per 1M 76.2
  5. 5 Muse Spark 1.3 (Max) $1.25 / $4.25 per 1M 75.8
  6. 6 GLM-5.3 (Max) $0.98 / $3.08 per 1M 74.8
  7. 7 Kimi K3 (Low) $0.16 / $22 per 1M 72
  8. 8 GLM-5.3-Flash $0.15 / $0.5 per 1M 71.5
  9. 9 Qwen3.8 Max 68.9
  10. 10 DeepSeek V4 Pro 0813 (Max) $0.12 / $4 per 1M 68.8
  11. 11 Qwen3.8 27B (Xhigh) $0.04 / $2.3 per 1M 68.1
  12. 12 MiniMax-M3 $0.22 / $0.88 per 1M 58.6
  13. 13 Inkling Small $0 / $0 per 1M 52.9
  14. 14 Solar Pro 4 $0.15 / $0.6 per 1M 52.7
  15. 15 Inkling (Xhigh) $0 / $0 per 1M 52.1
  16. 16 Nemotron 3 Ultra 550B A55B (Reasoning) $0.4 / $1.76 per 1M 49.3
  17. 17 Gemini 3.5 Flash-Lite $0.54 / $4.5 per 1M 49.3
  18. 18 Muse Glimmer (High) $0.3 / $1.2 per 1M 49
  19. 19 Command A+ $0.3 / $1.5 per 1M 27.8
  20. 20 Nemotron 3.5 Lightning $0.05 / $0.2 per 1M 26.8
  21. 21 Mistral Small 4 (Reasoning) $0.17 / $0.66 per 1M 26.6

Source: Artificial Analysis (artificialanalysis.ai) via OpenRouter (openrouter.ai/rankings).

τ²-Bench Airline

AI ทำหน้าที่พนักงานสายการบิน คุยหลายรอบ เรียก tool และต้องทำตามนโยบายเคร่งครัด · % งานที่ทำสำเร็จMulti-turn customer-service agents calling tools under strict policy · % of tasks solved

  1. 1 OpenAI: GPT-5.1 $0.393 / task · 100 tasks 80%
  2. 2 Anthropic: Claude Fable 5 $0.883 / task · 600 tasks 79.5%
  3. 3 inclusionAI: Ling 3.1 Flash $0.031 / task · 50 tasks 79.3%
  4. 4 Anthropic: Claude Fable 5.1 $0.716 / task · 200 tasks 79.2%
  5. 5 Anthropic: Claude Opus 5 $0.477 / task · 600 tasks 78.8%
  6. 6 Anthropic: Claude Sonnet 5.5 $0.159 / task · 250 tasks 78.8%
  7. 7 Google: Gemini 4 Argon $0.498 / task · 50 tasks 78.7%
  8. 8 Qwen: Qwen3.5 397B A17B $0.191 / task · 400 tasks 78.3%
  9. 9 Google: Gemini 3.5 Flash Lite $0.074 / task · 100 tasks 78.3%
  10. 10 Google: Gemini 3.7 Flash $0.150 / task · 146 tasks 78.3%
  11. 11 Xiaomi: MiMo-V2.6-Flash $0.015 / task · 250 tasks 77.9%
  12. 12 Qwen: Qwen3.8 27B $0.066 / task · 500 tasks 77.5%
  13. 13 DeepSeek: DeepSeek V4 Pro 0813 $0.101 / task · 550 tasks 77.5%
  14. 14 Qwen: Qwen3.5-122B-A10B $0.172 / task · 550 tasks 77.3%
  15. 15 Z.ai: GLM 5.3 $0.057 / task · 450 tasks 77.3%
  16. 16 Google: Gemini 3.1 Pro Preview $0.351 / task · 100 tasks 77%
  17. 17 DeepSeek: DeepSeek V4 Pro 0423 $0.050 / task · 400 tasks 76.8%
  18. 18 Xiaomi: MiMo-V2.6-Pro $0.068 / task · 300 tasks 76.8%
  19. 19 Anthropic: Claude Opus 4.7 $0.418 / task · 600 tasks 76.7%
  20. 20 Amazon: Nova Micro 1.0 $1.27 / task · 100 tasks 76.7%
  21. 21 Anthropic: Claude Opus 4.8 $0.494 / task · 600 tasks 76.6%
  22. 22 Anthropic: Claude Opus 4.5 $0.537 / task · 600 tasks 76.5%
  23. 23 Qwen: Qwen3.6 27B $0.141 / task · 550 tasks 76.2%
  24. 24 Tencent: Hy4 preview $0.114 / task · 250 tasks 76%
  25. 25 GLM 5.3 $0.016 / task · 50 tasks 76%

Agentic IndexAgentic Index

คะแนนการทำงานหลายขั้นตอนด้วยตัวเองจาก Artificial AnalysisAutonomous multi-step task score from Artificial Analysis

  1. 1 Claude Fable 5.1 (Xhigh, Default Fallback) 57.2
  2. 2 Qwen3.8 Max (0902) 56
  3. 3 Muse Spark 1.3 (Max) 55.5
  4. 4 GLM-5.3 (Max) 53.1
  5. 5 GPT-6 Astra (Max) 51
  6. 6 GLM-5.3-Flash 50.9
  7. 7 Qwen3.8 Max 49.9
  8. 8 Qwen3.8 27B (Xhigh) 45.8
  9. 9 DeepSeek V4 Pro 0813 (Max) 41.3
  10. 10 Gemini 3.8 Flash (High) 40.2
  11. 11 MiniMax-M3 29.5
  12. 12 Kimi K3 (Low) 29.3
  13. 13 Inkling Small 23.5
  14. 14 Inkling (Xhigh) 22.5
  15. 15 Nemotron 3 Ultra 550B A55B (Reasoning) 20.1

Source: OpenRouter evals (openrouter.ai) via OpenRouter (openrouter.ai/rankings). · Source: Artificial Analysis (artificialanalysis.ai) via OpenRouter (openrouter.ai/rankings).

GPQA Diamond

คำถามวิทยาศาสตร์ระดับบัณฑิตศึกษาที่ค้นหาคำตอบตรง ๆ ไม่ได้ ต้องใช้เหตุผล · % ตอบถูกGraduate-level science questions that resist lookup · % correct

  1. 1 Google: Gemini 4 Argon $0.095 / question 96.5%
  2. 2 Gemini 4 Argon $0.029 / question 96%
  3. 3 Google: Gemini 3.8 Flash $0.068 / question 95.6%
  4. 4 OpenAI: GPT-6 Astra Pro $0.333 / question 95.5%
  5. 5 Fugu Ultra $1.15 / question 94.6%
  6. 6 Google: Gemini 3.1 Pro Preview $0.202 / question 94.5%
  7. 7 OpenAI: GPT-6 Astra $0.113 / question 94.4%
  8. 8 OpenAI: GPT-6.1 Sol $0.021 / question 94.4%
  9. 9 OpenAI: GPT-5.6 Sol Pro $0.151 / question 94.1%
  10. 10 OpenAI: GPT-5.5 $0.313 / question 93.9%
  11. 11 Google: Gemini 3.7 Flash $0.029 / question 93.9%
  12. 12 Gemini 3.8 Flash $0.023 / question 93.4%
  13. 13 Fugu Ultra V2 $0.503 / question 93.3%
  14. 14 TypeSafe: Jev Router $0.035 / question 93.3%
  15. 15 SpaceXAI: Grok 4.6 $0.123 / question 92.8%
  16. 16 Anthropic: Claude Sonnet 5.5 $0.025 / question 92.8%
  17. 17 Google: Gemini 3.6 Flash $0.044 / question 92.8%
  18. 18 OpenAI: GPT-5.6 Sol $0.039 / question 92.4%
  19. 19 Pareto 26.10 Preview $0.003 / question 92.4%
  20. 20 Google: Gemini 3.5 Flash $0.144 / question 92.3%
  21. 21 OpenAI: GPT-6 Sol $0.025 / question 91.9%
  22. 22 MoonshotAI: Kimi K3 $0.132 / question 91.9%
  23. 23 GPT-6.1-Terra $0.008 / question 91.4%
  24. 24 OpenAI: GPT-6.1 Terra $0.011 / question 91.1%
  25. 25 GLM 5.3 Flash $0.005 / question 90.9%

Source: OpenRouter evals (openrouter.ai) via OpenRouter (openrouter.ai/rankings). · as of 31 Oct 2026

BrowseComp

หาข้อมูลที่หายากบนเว็บจริง ต้องค้นหลายขั้นHard-to-find facts on the live web

  1. 1 Claude Opus 5 · parallel $2.36 / task · 141s 89.2%
  2. 2 GPT-5.6-Sol · perplexity $0.500 / task · 116s 82.4%
  3. 3 Deepseek V4 Flash 0731 · perplexity $0.080 / task · 135s 77%
  4. 4 GPT-5.6-Luna · perplexity $0.100 / task · 111s 74%

WideSearch

เติมข้อมูลลงตารางทั้งตารางFill an entire table from the web

  1. 1 GPT-5.6-Sol · perplexity $0.830 / task · 136s 84%
  2. 2 GPT-5.6-Luna · perplexity $0.170 / task · 165s 83.4%
  3. 3 Deepseek V4 Flash 0731 · perplexity $0.100 / task · 119s 77.1%
  4. 4 Claude Opus 5 · parallel $1.36 / task · 218s 67.3%

DeepSearchQA

คำตอบเป็นรายการ ต้องค้นให้ครบList answers, scored for completeness

  1. 1 Claude Opus 5 · parallel $0.790 / task · 144s 88.7%
  2. 2 GPT-5.6-Sol · parallel $0.850 / task · 131s 88.4%
  3. 3 GPT-5.6-Luna · parallel $0.110 / task · 100s 87.5%
  4. 4 Deepseek V4 Flash 0731 · parallel $0.090 / task · 99s 81.7%

Humanity's Last Exam

คำถามระดับผู้เชี่ยวชาญ ตอบโดยค้นเว็บExpert questions answered with live search

  1. 1 GPT-5.6-Sol · parallel $0.710 / task · 85s 71%
  2. 2 Claude Opus 5 · parallel $0.410 / task · 125s 67.3%
  3. 3 Auto Router · auto $0.080 / task · 123s 52.5%

Source: OpenRouter search evals (openrouter.ai/benchmarks) · as of 16 Sept 2026

Design Arena

คนโหวตเทียบผลงานออกแบบจากโมเดลแบบไม่เห็นชื่อ แล้วคิดคะแนน Elo · เลือกหมวดPeople vote on blind side-by-side design outputs, scored as Elo · pick a category

  1. 1 Muse Wonder (max) win 62.6% · 0 battles 1370
  2. 2 Muse Spark 1.3 (xhigh) win 58.2% · 0 battles 1355
  3. 3 Muse Spark 1.3 Max win 56.3% · 0 battles 1344
  4. 4 Kimi K3 win 58.6% · 0 battles 1334
  5. 5 Kimi K3.1 Pro win 58.1% · 0 battles 1326
  6. 6 Claude Opus 5.5 win 55.1% · 0 battles 1325
  7. 7 Claude Sonnet 5.5 win 54.7% · 0 battles 1319
  8. 8 Gemini 4 Argon win 55.3% · 0 battles 1315
  9. 9 GPT-6.1 Sol win 53.5% · 0 battles 1314
  10. 10 Muse Spark Open (xhigh) win 56.4% · 0 battles 1312
  11. 11 Muse Spark 1.2 win 52.3% · 0 battles 1311
  12. 12 Claude Fable 5.1 win 55.1% · 0 battles 1308
  13. 13 Claude Opus 5 win 56% · 0 battles 1308
  14. 14 MiMo-V2.6-Pro win 50.8% · 0 battles 1306
  15. 15 Gemini 3.7 Flash win 54% · 0 battles 1302
  1. 1 Kimi K3.1 Pro win 66.1% · 0 battles 1394
  2. 2 Muse Wonder (max) win 63.4% · 0 battles 1379
  3. 3 Claude Opus 5.5 win 63.5% · 0 battles 1378
  4. 4 Claude Sonnet 5.5 win 63.3% · 0 battles 1375
  5. 5 Muse Spark 1.3 (xhigh) win 57.8% · 0 battles 1346
  6. 6 Claude Opus 5 win 60.9% · 0 battles 1345
  7. 7 Muse Spark 1.3 Max win 57.9% · 0 battles 1345
  8. 8 GPT-6.1 Sol win 58.6% · 0 battles 1343
  9. 9 Qwen3.8 Max win 57.9% · 0 battles 1339
  10. 10 Kimi K3 win 59.9% · 0 battles 1338
  11. 11 MiMo-V2.6-Pro win 58.7% · 0 battles 1335
  12. 12 Muse Spark Open (xhigh) win 55.1% · 0 battles 1335
  13. 13 Claude Fable 5.1 win 59.6% · 0 battles 1333
  14. 14 Gemini 4 Argon win 57.8% · 0 battles 1332
  15. 15 Gemini 3.8 Flash win 53.6% · 0 battles 1314
  1. 1 Claude Opus 5.5 win 64.4% · 0 battles 1394
  2. 2 GLM 5.1 win 67% · 5,939 battles 1366
  3. 3 Claude Sonnet 5.5 win 64.1% · 0 battles 1361
  4. 4 Kimi K3 win 63.2% · 0 battles 1351
  5. 5 Claude Fable 5.1 win 61.8% · 0 battles 1344
  6. 6 Muse Spark 1.3 (xhigh) win 58.2% · 0 battles 1344
  7. 7 Gemini 4 Argon win 61.5% · 0 battles 1335
  8. 8 MiMo-V2.6-Pro win 58.4% · 0 battles 1335
  9. 9 Claude Opus 5 win 58.5% · 0 battles 1331
  10. 10 Muse Spark 1.2 win 57.2% · 0 battles 1331
  11. 11 Gemini 3.7 Flash win 55.5% · 0 battles 1318
  12. 12 Muse Spark 1.3 Max win 52.5% · 0 battles 1311
  13. 13 Muse Wonder (max) win 55.4% · 0 battles 1306
  14. 14 Claude Fable 5 win 57.5% · 0 battles 1304
  15. 15 Kimi K3.1 Pro win 55.7% · 0 battles 1300
  1. 1 Claude Opus 5.5 win 64% · 0 battles 1434
  2. 2 Muse Wonder (max) win 61.6% · 0 battles 1415
  3. 3 GPT-6.1 Sol win 60.2% · 0 battles 1414
  4. 4 Kimi K3.1 Pro win 67.1% · 0 battles 1412
  5. 5 Claude Sonnet 5.5 win 61.9% · 0 battles 1408
  6. 6 Claude Fable 5.1 win 60.1% · 0 battles 1373
  7. 7 Gemini 4 Argon win 53.9% · 0 battles 1367
  8. 8 Kimi K3 win 53.9% · 0 battles 1349
  9. 9 Muse Spark 1.3 Max win 56% · 0 battles 1342
  10. 10 Claude Opus 5 win 58.6% · 0 battles 1340
  11. 11 Claude Fable 5 win 60.8% · 0 battles 1335
  12. 12 Muse Spark 1.3 (xhigh) win 54.1% · 0 battles 1331
  13. 13 Claude Haiku 5.5 win 54.7% · 0 battles 1320
  14. 14 Gemini 3.7 Flash win 53.6% · 0 battles 1318
  15. 15 Grok 4.7 win 53% · 0 battles 1316
  1. 1 Claude Opus 5.5 win 64.5% · 0 battles 1471
  2. 2 Muse Wonder (max) win 60.3% · 0 battles 1444
  3. 3 Kimi K3.1 Pro win 71.3% · 0 battles 1440
  4. 4 GPT-6.1 Sol win 58.4% · 0 battles 1437
  5. 5 Claude Fable 5.1 win 63.4% · 0 battles 1403
  6. 6 Claude Sonnet 5.5 win 51.5% · 0 battles 1399
  7. 7 Muse Spark 1.3 Max win 63.5% · 0 battles 1393
  8. 8 Kimi K3 win 64.3% · 0 battles 1382
  9. 9 Gemini 4 Argon win 50.3% · 0 battles 1370
  10. 10 DeepSeek-V4.1-Flash (1017) win 47% · 0 battles 1360
  11. 11 Muse Spark 1.3 (xhigh) win 56.5% · 0 battles 1360
  12. 12 Claude Haiku 5.5 win 57.3% · 0 battles 1338
  13. 13 GLM 5.1 win 62.6% · 5,939 battles 1336
  14. 14 Qwen3.8 Max win 56.3% · 0 battles 1334
  15. 15 Claude Opus 5 win 59.7% · 0 battles 1333
  1. 1 Gemini 4 Argon win 60.7% · 0 battles 1329
  2. 2 Kimi K3.1 Pro win 62.4% · 0 battles 1329
  3. 3 Claude Opus 5.5 win 52.5% · 0 battles 1323
  4. 4 Muse Spark 1.3 Max win 55.2% · 0 battles 1322
  5. 5 Claude Opus 5 win 58.2% · 0 battles 1307
  6. 6 Kimi K3 win 60.1% · 0 battles 1305
  7. 7 MiMo-V2.6-Pro win 54.8% · 0 battles 1301
  8. 8 Muse Spark 1.3 (xhigh) win 55.6% · 0 battles 1299
  9. 9 Claude Fable 5 win 62.6% · 0 battles 1298
  10. 10 Claude Fable 5.1 win 56.4% · 0 battles 1296
  11. 11 Gemini 3.8 Flash win 54.8% · 0 battles 1295
  12. 12 Gemini 3.1 Pro Preview win 67.8% · 0 battles 1283
  13. 13 Gemini 3.7 Flash win 53.8% · 0 battles 1283
  14. 14 Muse Spark 1.2 win 54.6% · 0 battles 1279
  15. 15 GLM-5.3-Flash win 55.4% · 0 battles 1279
  1. 1 Muse Wonder (max) win 61.8% · 0 battles 1386
  2. 2 Claude Opus 5.5 win 60.6% · 0 battles 1385
  3. 3 GPT-6.1 Sol win 56.4% · 0 battles 1359
  4. 4 Kimi K3.1 Pro win 61.1% · 0 battles 1354
  5. 5 Kimi K3 win 60% · 0 battles 1352
  6. 6 Claude Sonnet 5.5 win 56.4% · 0 battles 1350
  7. 7 Muse Spark 1.3 (xhigh) win 57% · 0 battles 1350
  8. 8 Muse Spark 1.3 Max win 56.8% · 0 battles 1349
  9. 9 Claude Fable 5.1 win 57.4% · 0 battles 1333
  10. 10 Gemini 4 Argon win 53.8% · 0 battles 1327
  11. 11 Claude Opus 5 win 57.4% · 0 battles 1323
  12. 12 MiMo-V2.6-Pro win 52.8% · 0 battles 1316
  13. 13 Muse Spark 1.2 win 52.1% · 0 battles 1311
  14. 14 Claude Fable 5 win 57.6% · 0 battles 1309
  15. 15 Muse Spark Open (xhigh) win 54.9% · 0 battles 1307
  1. 1 Riverflow 2.5 Pro win 63.7% · 0 battles 1297
  2. 2 MAI-Image-2.6 win 58.4% · 0 battles 1292
  3. 3 MAI-Image-2.6 Flash win 56% · 0 battles 1268
  4. 4 Gemini 3.1 Flash Image Gen 2K (Nano Banana 2) win 62.9% · 0 battles 1263
  5. 5 Gemini 3.1 Flash Image Gen (Nano Banana 2) win 61.1% · 0 battles 1248
  6. 6 MAI-Image-2.5 win 53.6% · 0 battles 1248
  7. 7 Gemini 3 Pro Image Gen 2K (Nano Banana Pro) win 61% · 0 battles 1241
  8. 8 FLUX.2 [flex] win 54% · 0 battles 1209
  9. 9 GPT-Image-1 Mini win 51.4% · 0 battles 1209
  10. 10 GPT-Image-1 win 52.7% · 0 battles 1202
  11. 11 FLUX.2 [pro] win 48.8% · 0 battles 1192
  12. 12 Gemini 3 Pro Image Preview win 54.2% · 0 battles 1181
  13. 13 Gemini 2.5 Flash Image (Nano Banana) win 51.4% · 0 battles 1169
  14. 14 Gemini 2.5 Flash Image Gen (Nano Banana) win 51.2% · 0 battles 1165
  15. 15 FLUX.2 Klein 4B win 38% · 0 battles 1060
  1. 1 Riverflow 2.5 Pro win 65.8% · 0 battles 1350
  2. 2 MAI-Image-2.6 win 61.6% · 0 battles 1317
  3. 3 MAI-Image-2.6 Flash win 60.6% · 0 battles 1305
  4. 4 Gemini 3.1 Flash Image Gen 2K (Nano Banana 2) win 65.1% · 0 battles 1274
  5. 5 Gemini 3.1 Flash Image Gen (Nano Banana 2) win 64.4% · 0 battles 1266
  6. 6 Gemini 3 Pro Image Gen 2K (Nano Banana Pro) win 62.1% · 0 battles 1241
  7. 7 MAI-Image-2.5 win 54.6% · 0 battles 1227
  8. 8 Gemini 3 Pro Image Preview win 60.4% · 0 battles 1212
  9. 9 FLUX.2 [flex] win 55.1% · 0 battles 1196
  10. 10 FLUX.2 [pro] win 49.1% · 0 battles 1192
  11. 11 GPT-Image-1 win 53.9% · 0 battles 1188
  12. 12 Gemini 2.5 Flash Image Gen (Nano Banana) win 55.6% · 0 battles 1183
  13. 13 GPT-Image-1 Mini win 51% · 0 battles 1181
  14. 14 Gemini 2.5 Flash Image (Nano Banana) win 53.4% · 0 battles 1176
  15. 15 FLUX.2 Klein 4B Distilled win 38.8% · 0 battles 1060

Source: Design Arena (www.designarena.ai) via OpenRouter (openrouter.ai/rankings). · as of 31 Oct 2026

โมเดลเปิดตัวล่าสุดNewest models

โมเดลที่เพิ่งเปิดให้ใช้งานผ่าน OpenRouter พร้อมราคาและ contextRecently released on OpenRouter, with pricing and context length

โมเดลModelเปิดตัวReleasedContextInput $/1MOutput $/1M
Particle AI: Particle-1particle-ai/particle-1 29 Oct 2026 1.0M $0.05 $0.5
OpenAI: GPT-6.1 Terra Proopenai/gpt-6.1-terra-pro 27 Oct 2026 1.1M $1 $5
OpenAI: GPT-6.1 Terra Pro (batch)openai/gpt-6.1-terra-pro:batch 27 Oct 2026 1.1M $0.5 $2.5
OpenAI: GPT-6.1 Terraopenai/gpt-6.1-terra 27 Oct 2026 1.1M $1 $5
OpenAI: GPT-6.1 Terra (batch)openai/gpt-6.1-terra:batch 27 Oct 2026 1.1M $0.5 $2.5
MoonshotAI: Kimi K3.1 Promoonshotai/kimi-k3.1-pro 24 Oct 2026 1.0M $2 $10
Recursive: Flint (free)recursive/flint-v1.0:free 22 Oct 2026 262K free free
MiniMax: MiniMax M3.1 Flashminimax/minimax-m3.1-flash 19 Oct 2026 1.0M $0.08 $0.3
DeepSeek: DeepSeek V4.1 Flash 1017deepseek/deepseek-v4.1-flash-1017 17 Oct 2026 1.0M $0.15 $0.6
Elorian: Glimpseelorian/glimpse 17 Oct 2026 262K $0.6 $3
XingChen: Xing4.0 29B A4B (free)xingchen-agi/xing4.0-29b-a4b:free 17 Oct 2026 262K free free
Reflection: Beam 1.0 501B A23Breflection/beam-1.0-501b-a23b 16 Oct 2026 1.0M $0.25 $1
Inception: Mercury 3inception/mercury-3 16 Oct 2026 1M $0.1 $0.35
PrismML: Ternary Bonsai 2 9Bprism-ml/ternary-bonsai-2-9b 16 Oct 2026 262K $0.08 $0.13
Moondream: Moondream 3 Previewmoondream/moondream3-preview 16 Oct 2026 32K $0.3 $2.5
Moondream: Moondream 3.1moondream/moondream3.1-9b-a2b 16 Oct 2026 32K $0.3 $2.5
Meta: Muse Spark Openmeta/muse-spark-open 16 Oct 2026 1.0M $1.25 $4.25
NVIDIA: Nemotron 3.5 Super VLnvidia/nemotron-3.5-super-vl 15 Oct 2026 1.0M $0.15 $0.5
NVIDIA: Nemotron 3.5 Super VL (free)nvidia/nemotron-3.5-super-vl:free 15 Oct 2026 1M free free
Pareto 26.10unbiased/pareto-26.10 15 Oct 2026 1.0M $0.8 $3.2

Source: OpenRouter models API (openrouter.ai/models) · 452 models

แหล่งข้อมูลและวิธีวัดData sources & methodology

ตัวเลขทั้งหมดดึงผ่าน OpenRouter Benchmarks API และ Models API ทุกคืน (01:00 น.) — เราไม่ได้ปรับแก้คะแนนAll numbers are pulled nightly (01:00 Bangkok) through the OpenRouter Benchmarks & Models APIs — we do not adjust any scores.

Artificial Analysis

บริษัทวิจัยอิสระที่รันชุดทดสอบมาตรฐานกับทุกโมเดลด้วยเงื่อนไขเดียวกัน แล้วสรุปเป็นดัชนี Intelligence, Coding และ Agentic พร้อมราคาต่อ tokenIndependent lab that runs the same standard eval suites on every model and publishes Intelligence, Coding and Agentic indexes plus token pricing.

ใช้ในแท็บUsed in: ฉลาดสุด · เขียนโค้ด · AgentIntelligence · Coding · Agents · 36 models · as of 30 Oct 2026

OpenRouter Evals

OpenRouter รันการทดสอบเองแบบทำซ้ำได้ ทุกคะแนนลิงก์ถึง config ค่าใช้จ่าย และ telemetry: GPQA Diamond (วิทยาศาสตร์), τ²-Bench Airline (agent เรียก tool) และชุดค้นเว็บ BrowseComp / WideSearch / DeepSearchQA / HLEReproducible evals run by OpenRouter, each linked to its config, cost and telemetry: GPQA Diamond, τ²-Bench Airline, and the BrowseComp / WideSearch / DeepSearchQA / HLE search suite.

ใช้ในแท็บUsed in: Agent · GPQA · AI ค้นเว็บWeb search · 318 results · as of 31 Oct 2026

Design Arena

ผู้ใช้จริงโหวตเทียบผลงาน (เว็บไซต์ UI กราฟ เกม 3D โลโก้ ฯลฯ) จากสองโมเดลโดยไม่เห็นชื่อ แล้วคำนวณคะแนน Elo และอัตราชนะReal users vote on blind side-by-side outputs (websites, UI, charts, games, 3D, logos…), aggregated into Elo ratings and win rates.

ใช้ในแท็บUsed in: ออกแบบ UIDesign · 14 categories · as of 31 Oct 2026

OpenRouter Models API

แคตตาล็อกโมเดลที่เปิดให้ใช้งานผ่าน API พร้อมวันเปิดตัว context window และราคา input / output ต่อ tokenCatalogue of models available via API with release date, context window and input / output token pricing.

ใช้ในแท็บUsed in: โมเดลใหม่ & ราคาNew & pricing · 452 models

ข้อมูลภายใต้สัญญาอนุญาต CC BY 4.0 ของ OpenRouter Rankings — แสดงเครดิตแหล่งต้นทางตามที่ API กำหนด · Benchmark วัดเฉพาะงานที่ทดสอบ ควรทดลองกับงานจริงของคุณก่อนเลือกใช้Rankings data is licensed CC BY 4.0 by OpenRouter — credited to each original source as the API specifies · Benchmarks cover only the tasks they test; try models on your own workload before choosing.

ไปที่เพจ FacebookOpen Facebook page">