German Language LLM Index
a PeerBench project

Fresh evidence

New results

Completed benchmark runs, newest first. Every entry links to a permanent static result page.

Published runs

202 canonical results across 46 models.

Follow via RSS →
  1. Gemini 3.5 Flash-Lite · GermEval NER (DE)

    Google Micro-F1 (no MISC) provider-internal

    85.4%
  2. Gemini 3.5 Flash-Lite · INCLUDE (DE)

    Google Exact match provider-internal

    85.0%
  3. Gemini 3.5 Flash-Lite · MuSR (DE)

    Google Accuracy provider-internal

    82.3%
  4. Gemini 3.5 Flash-Lite · SB10K (DE)

    Google MCC provider-internal

    62.5%
  5. Gemini 3.5 Flash-Lite · ScaLA (DE)

    Google MCC provider-internal

    75.1%
  6. GLM-5.3 Flash · GermEval NER (DE)

    Z.ai Micro-F1 (no MISC) fp8

    86.4%
  7. 86.9%
  8. 81.9%
  9. 63.6%
  10. 77.2%
  11. Tencent HY3 · GermEval NER (DE)

    Tencent Micro-F1 (no MISC) fp8

    76.1%
  12. 83.2%
  13. 85.0%
  14. 62.4%
  15. 69.7%
  16. 82.1%
  17. 86.9%
  18. 80.1%
  19. 62.4%
  20. 74.0%
  21. Gemini 3.1 Pro · MMLU-ProX (DE)

    Google Exact match provider-internal

    89.2%
  22. 86.0%
  23. 86.1%
  24. 86.9%
  25. 82.7%
  26. 83.2%
  27. 82.3%
  28. 84.4%
  29. 61.0%
  30. 71.7%
  31. 81.5%
  32. 87.9%
  33. 79.9%
  34. 86.1%
  35. 62.6%
  36. 66.9%
  37. 81.6%
  38. 86.9%
  39. 80.0%
  40. 62.7%
  41. 66.3%
  42. 80.7%
  43. Gemini 3.7 Flash · GermEval NER (DE)

    Google Micro-F1 (no MISC) provider-internal

    86.9%
  44. Gemini 3.7 Flash · INCLUDE (DE)

    Google Exact match provider-internal

    93.5%
  45. Gemini 3.7 Flash · MMLU-ProX (DE)

    Google Exact match provider-internal

    88.2%
  46. Gemini 3.7 Flash · MuSR (DE)

    Google Accuracy provider-internal

    88.6%
  47. Gemini 3.7 Flash · SB10K (DE)

    Google MCC provider-internal

    67.5%
  48. Gemini 3.7 Flash · ScaLA (DE)

    Google MCC provider-internal

    79.1%
  49. DeepSeek V4 Pro 0813 · GermEval NER (DE)

    DeepSeek Micro-F1 (no MISC) provider-internal

    81.7%
  50. DeepSeek V4 Pro 0813 · INCLUDE (DE)

    DeepSeek Exact match provider-internal

    87.9%
  51. DeepSeek V4 Pro 0813 · MuSR (DE)

    DeepSeek Accuracy provider-internal

    82.3%
  52. 60.8%
  53. 75.6%
  54. GPT-5.6 Luna · GermEval NER (DE)

    OpenAI Micro-F1 (no MISC) provider-internal

    81.9%
  55. GPT-5.6 Luna · INCLUDE (DE)

    OpenAI Exact match provider-internal

    86.0%
  56. GPT-5.6 Luna · MuSR (DE)

    OpenAI Accuracy provider-internal

    75.8%
  57. GPT-5.6 Luna · SB10K (DE)

    OpenAI MCC provider-internal

    61.9%
  58. GPT-5.6 Luna · ScaLA (DE)

    OpenAI MCC provider-internal

    72.2%
  59. GPT-5.6 Terra · INCLUDE (DE)

    OpenAI Exact match provider-internal

    88.8%
  60. GPT-5.6 Terra · MuSR (DE)

    OpenAI Accuracy provider-internal

    78.7%
  61. GPT-5.6 Terra · GermEval NER (DE)

    OpenAI Micro-F1 (no MISC) provider-internal

    83.3%
  62. GPT-5.6 Terra · SB10K (DE)

    OpenAI MCC provider-internal

    62.7%
  63. GPT-5.6 Terra · ScaLA (DE)

    OpenAI MCC provider-internal

    75.1%
  64. 80.3%
  65. 80.7%
  66. 85.3%
  67. 61.9%
  68. 74.0%
  69. 82.0%
  70. 61.2%
  71. 76.1%
  72. MiMo V2.5 · GermEval NER (DE)

    Xiaomi Micro-F1 (no MISC) fp8

    80.8%
  73. 61.2%
  74. 53.4%
  75. Gemini 3.5 Flash-Lite · MMLU-ProX (DE)

    Google Exact match provider-internal

    80.2%
  76. Gemini 3.6 Flash · GermEval NER (DE)

    Google Micro-F1 (no MISC) provider-internal

    86.2%
  77. Gemini 3.6 Flash · INCLUDE (DE)

    Google Exact match provider-internal

    74.8%
  78. Gemini 3.6 Flash · MMLU-ProX (DE)

    Google Exact match provider-internal

    86.0%
  79. Gemini 3.6 Flash · MuSR (DE)

    Google Accuracy provider-internal

    80.3%
  80. Gemini 3.6 Flash · SB10K (DE)

    Google MCC provider-internal

    64.5%
  81. Gemini 3.6 Flash · ScaLA (DE)

    Google MCC provider-internal

    77.6%
  82. Claude Sonnet 5 · INCLUDE (DE)

    Anthropic Exact match provider-internal

    72.7%
  83. Claude Fable 5 · GermEval NER (DE)

    Anthropic Micro-F1 (no MISC) provider-internal

    87.7%
  84. Claude Fable 5 · SB10K (DE)

    Anthropic MCC provider-internal

    67.3%
  85. Claude Fable 5 · ScaLA (DE)

    Anthropic MCC provider-internal

    83.6%
  86. Claude Fable 5 · INCLUDE (DE)

    Anthropic Exact match provider-internal

    72.7%
  87. Claude Fable 5 · MuSR (DE)

    Anthropic Accuracy provider-internal

    89.5%
  88. GPT-5.5 · MMLU-ProX (DE)

    OpenAI Exact match provider-internal

    87.2%
  89. Gemini 3 Flash Preview · GermEval NER (DE)

    Google Micro-F1 (no MISC) provider-internal

    84.7%
  90. Gemini 3 Flash Preview · INCLUDE (DE)

    Google Exact match provider-internal

    71.9%
  91. Gemini 3 Flash Preview · MMLU-ProX (DE)

    Google Exact match provider-internal

    86.3%
  92. Gemini 3 Flash Preview · MuSR (DE)

    Google Accuracy provider-internal

    79.3%
  93. 63.6%
  94. 77.0%
  95. DeepSeek V4 Pro · MuSR (DE)

    DeepSeek Accuracy provider-internal

    85.3%
  96. DeepSeek V4 Flash · MMLU-ProX (DE)

    DeepSeek Exact match provider-internal

    74.9%
  97. DeepSeek V4 Pro · MMLU-ProX (DE)

    DeepSeek Exact match provider-internal

    80.8%
  98. Claude Opus 4.8 · GermEval NER (DE)

    Anthropic Micro-F1 (no MISC) provider-internal

    85.8%
  99. Claude Opus 4.8 · MuSR (DE)

    Anthropic Accuracy provider-internal

    86.3%
  100. Claude Opus 4.8 · SB10K (DE)

    Anthropic MCC provider-internal

    64.5%
  101. Claude Opus 4.8 · ScaLA (DE)

    Anthropic MCC provider-internal

    82.6%
  102. GPT-5.5 · GermEval NER (DE)

    OpenAI Micro-F1 (no MISC) provider-internal

    85.9%
  103. GPT-5.5 · MuSR (DE)

    OpenAI Accuracy provider-internal

    86.3%
  104. GPT-5.5 · SB10K (DE)

    OpenAI MCC provider-internal

    62.5%
  105. GPT-5.5 · ScaLA (DE)

    OpenAI MCC provider-internal

    79.7%
  106. Claude Opus 4.8 · INCLUDE (DE)

    Anthropic Exact match provider-internal

    71.9%
  107. GPT-5.5 · INCLUDE (DE)

    OpenAI Exact match provider-internal

    74.8%
  108. Gemma 4 12B · GermEval NER (DE)

    Google Micro-F1 (no MISC) bf16

    79.6%
  109. Gemma 4 12B · INCLUDE (DE)

    Google Exact match bf16

    69.1%
  110. 73.1%
  111. 81.6%
  112. 62.2%
  113. 64.1%
  114. Gemma 4 31B · GermEval NER (DE)

    Google Micro-F1 (no MISC) bf16

    82.6%
  115. Gemma 4 31B · INCLUDE (DE)

    Google Exact match bf16

    68.3%
  116. 82.1%
  117. 83.5%
  118. 61.0%
  119. 71.0%
  120. Qwen3 14B · GermEval NER (DE)

    Alibaba Micro-F1 (no MISC) bf16

    73.0%
  121. Qwen3 14B · INCLUDE (DE)

    Alibaba Exact match bf16

    63.3%
  122. Qwen3 14B · MMLU-ProX (DE)

    Alibaba Exact match bf16

    64.1%
  123. Qwen3 14B · MuSR (DE)

    Alibaba Accuracy bf16

    69.7%
  124. 56.1%
  125. 64.0%
  126. Qwen3.5 9B · GermEval NER (DE)

    Alibaba Micro-F1 (no MISC) bf16

    72.6%
  127. Qwen3.5 9B · INCLUDE (DE)

    Alibaba Exact match bf16

    64.7%
  128. 73.4%
  129. 80.9%
  130. 55.7%
  131. 83.5%
  132. Gemini 3.1 Flash-Lite · INCLUDE (DE)

    Google Exact match provider-internal

    71.9%
  133. Gemini 3.1 Pro · INCLUDE (DE)

    Google Exact match provider-internal

    77.7%
  134. Gemini 3.5 Flash · INCLUDE (DE)

    Google Exact match provider-internal

    74.1%
  135. 83.7%
  136. GLM-5.1 · MuSR (DE)

    Z.ai Accuracy fp8

    81.4%
  137. 80.9%
  138. 62.4%
  139. 80.2%
  140. 62.4%
  141. 74.9%
  142. 82.2%
  143. 60.0%
  144. 73.5%
  145. Gemini 2.5 Flash · GermEval NER (DE)

    Google Micro-F1 (no MISC) provider-internal

    81.8%
  146. Gemini 2.5 Flash · MuSR (DE)

    Google Accuracy provider-internal

    83.3%
  147. Gemini 2.5 Flash · SB10K (DE)

    Google MCC provider-internal

    61.6%
  148. Gemini 2.5 Flash · ScaLA (DE)

    Google MCC provider-internal

    77.1%
  149. Gemini 3.1 Flash-Lite · MuSR (DE)

    Google Accuracy provider-internal

    84.4%
  150. Gemini 3.1 Pro · GermEval NER (DE)

    Google Micro-F1 (no MISC) provider-internal

    87.3%
  151. Gemini 3.1 Pro · MuSR (DE)

    Google Accuracy provider-internal

    88.1%
  152. Gemini 3.1 Pro · SB10K (DE)

    Google MCC provider-internal

    70.1%
  153. Gemini 3.1 Pro · ScaLA (DE)

    Google MCC provider-internal

    80.4%
  154. Gemini 3.5 Flash · GermEval NER (DE)

    Google Micro-F1 (no MISC) provider-internal

    86.0%
  155. Gemini 3.5 Flash · MuSR (DE)

    Google Accuracy provider-internal

    84.2%
  156. Gemini 3.5 Flash · SB10K (DE)

    Google MCC provider-internal

    64.3%
  157. Gemini 3.5 Flash · ScaLA (DE)

    Google MCC provider-internal

    79.4%
  158. GLM-5.1 · GermEval NER (DE)

    Z.ai Micro-F1 (no MISC) fp8

    52.2%
  159. 23.1%
  160. 69.9%
  161. MiMo V2.5 Pro · GermEval NER (DE)

    Xiaomi Micro-F1 (no MISC) fp8

    81.9%
  162. 61.7%
  163. 74.8%
  164. Qwen3.6 35B-A3B · GermEval NER (DE)

    Alibaba Micro-F1 (no MISC) fp8

    79.9%
  165. 58.4%
  166. 64.2%
  167. Qwen3.7 Max · GermEval NER (DE)

    Alibaba Micro-F1 (no MISC) provider-internal

    82.1%
  168. Qwen3.7 Max · SB10K (DE)

    Alibaba MCC provider-internal

    63.2%
  169. Qwen3.7 Max · ScaLA (DE)

    Alibaba MCC provider-internal

    69.8%
  170. Tencent HY3-Preview · GermEval NER (DE)

    Tencent Micro-F1 (no MISC) provider-internal

    77.3%
  171. Tencent HY3-Preview · SB10K (DE)

    Tencent MCC provider-internal

    38.9%
  172. Tencent HY3-Preview · ScaLA (DE)

    Tencent MCC provider-internal

    62.5%
  173. 70.5%
  174. 70.5%
  175. Gemini 3.1 Flash-Lite · GermEval NER (DE)

    Google Micro-F1 (no MISC) provider-internal

    83.2%
  176. Gemini 3.1 Flash-Lite · SB10K (DE)

    Google MCC provider-internal

    60.4%
  177. Gemini 3.1 Flash-Lite · ScaLA (DE)

    Google MCC provider-internal

    74.2%
  178. Gemma 4 26B A4B · GermEval NER (DE)

    Google Micro-F1 (no MISC) provider-internal

    81.8%
  179. Gemma 4 26B A4B · SB10K (DE)

    Google MCC provider-internal

    15.2%
  180. Gemma 4 26B A4B · ScaLA (DE)

    Google MCC provider-internal

    66.3%
  181. GLM-5.1 · INCLUDE (DE)

    Z.ai Exact match fp8

    82.2%
  182. 69.8%
  183. 67.6%
  184. Qwen3.7 Max · INCLUDE (DE)

    Alibaba Exact match provider-internal

    71.2%
  185. 78.2%
  186. Claude Haiku 4.5 · MMLU-ProX (DE)

    Anthropic Exact match provider-internal

    75.3%
  187. Gemini 2.5 Flash-Lite · MMLU-ProX (DE)

    Google Exact match provider-internal

    71.2%
  188. Gemini 3.1 Flash-Lite · MMLU-ProX (DE)

    Google Exact match provider-internal

    82.2%
  189. 53.2%
  190. 64.7%
  191. 63.3%
  192. 58.3%
  193. Claude Haiku 4.5 · INCLUDE (DE)

    Anthropic Exact match provider-internal

    68.3%
  194. Gemini 2.5 Flash · INCLUDE (DE)

    Google Exact match provider-internal

    70.5%
  195. 68.3%
  196. Tencent HY3-Preview · INCLUDE (DE)

    Tencent Exact match provider-internal

    69.1%
  197. 80.0%
  198. 66.2%
  199. Tencent HY3-Preview · MMLU-ProX (DE)

    Tencent Exact match provider-internal

    73.2%
  200. Gemini 2.5 Flash · MMLU-ProX (DE)

    Google Exact match provider-internal

    79.6%
  201. Gemini 3.5 Flash · MMLU-ProX (DE)

    Google Exact match provider-internal

    86.5%
  202. Undated 72.7%