← Back to Test Cases
EM Excursion Investigation Report
Task CompletionEm Reportmedium
Cross-Model Comparison
| Model | Score | Latency | Tokens In | Tokens Out |
|---|---|---|---|---|
| Qwen3.8-Flash | N/A | N/A | - | - |
| GLM-5.3-Flash | N/A | N/A | - | - |
| GLM-5.3 | N/A | N/A | - | - |
| Muse Glimmer 30B | N/A | N/A | - | - |
| Muse Glimmer 30B | N/A | N/A | - | - |
| MiniMax M3 | N/A | N/A | - | - |
| Qwen3.8 Max (2.4T-A95B) | N/A | N/A | - | - |
| Qwen3.8 Max (2.4T-A95B) | N/A | N/A | - | - |
| Qwen3.8 Max (2.4T-A95B) | N/A | N/A | - | - |
| Qwen3.8-Flash | N/A | N/A | - | - |
| Qwen3.8-Flash | N/A | N/A | - | - |
| GLM-5.3 | N/A | N/A | - | - |
| GLM-5.3 | N/A | N/A | - | - |
| GLM-5.3-Flash | N/A | N/A | - | - |
| GLM-5.3-Flash | N/A | N/A | - | - |
| Hy4 Preview | N/A | N/A | - | - |
| MiniMax M3 | N/A | N/A | - | - |
| Hy4 Preview | N/A | N/A | - | - |
| Hy4 Preview | N/A | N/A | - | - |
| Muse Glimmer 30B | N/A | N/A | - | - |
| MiniMax M3 | N/A | N/A | - | - |
| Claude Opus 4.8 | 92.4% | 57.4s | 439 | 4,287 |
| Qwen3.6 27B | 91.5% | 86.2s | 340 | 5,299 |
| MiniMax M2.7 | 91.5% | 95.1s | 290 | 6,380 |
| Claude Opus 4.6 | 90.9% | 186.4s | 333 | 8,303 |
| DeepSeek V4 Pro | 89.8% | 49.3s | 282 | 2,973 |
| Claude Sonnet 4.6 | 89.6% | 152.4s | 333 | 8,388 |
| Qwen3.6 35B A3B | 89.5% | 32.6s | 340 | 5,285 |
| Claude Haiku 4.5 | 89.5% | 57.2s | 332 | 6,965 |
| Qwen3.5-397B-A17B | 88.6% | 59.2s | 340 | 4,168 |
| GLM-5.3 | 87.5% | 162.9s | 295 | 20,727 |
| Gemini 3.1 Pro | 87.3% | 46.8s | 322 | 3,903 |
| Hy4 Preview | 87.1% | 291.7s | 303 | 19,350 |
| Qwen3.8 Max (2.4T-A95B) | 86.6% | 367.8s | 378 | 13,875 |
| DeepSeek-V3.2 | 85.9% | 80.0s | 282 | 1,945 |
| Qwen3.8 Max (2.4T-A95B) | 85.0% | 349.2s | 378 | 13,812 |
| GPT-5.6 Terra | 84.8% | 66.3s | 290 | 4,454 |
| Qwen3.8 Max (2.4T-A95B) | 84.8% | 325.1s | 378 | 12,733 |
| GPT-5.6 Luna | 84.8% | 39.8s | 290 | 3,812 |
| Qwen3.8 27B | 84.6% | 158.2s | 378 | 12,024 |
| Qwen3.8-Flash | 84.3% | 186.3s | 378 | 11,124 |
| Qwen3.8 27B | 84.2% | 347.1s | 378 | 14,856 |
| GPT-5.6 Sol | 83.7% | 66.3s | 290 | 4,591 |
| GPT-5.6 Terra | 83.5% | 64.5s | 290 | 4,218 |
| GPT-5.6 Luna | 83.4% | 46.6s | 290 | 4,071 |
| DeepSeek-R1 | 83.4% | 54.2s | 282 | 1,965 |
| Qwen3.5-35B-A3B | 83.4% | 30.6s | 342 | 4,056 |
| Claude Sonnet 5 | 83.4% | 62.2s | 439 | 5,835 |
| Qwen3.8-Flash | 83.3% | 151.8s | 378 | 10,774 |
| Hy4 Preview | 83.2% | 214.8s | 303 | 15,354 |
| GLM-5.3-Flash | 83.1% | 303.2s | 295 | 12,130 |
| Muse Glimmer 30B | 83.0% | 42.3s | 303 | 2,819 |
| Claude Sonnet 5 | 83.0% | 56.4s | 439 | 4,861 |
| GLM-5.3-Flash | 82.3% | 377.3s | 295 | 11,221 |
| GLM-5.2 | 81.8% | 32.2s | 288 | 2,165 |
| MiniMax M3 | 81.5% | 64.1s | 444 | 6,136 |
| Gemini 3 Flash | 81.5% | 8.5s | 320 | 1,286 |
| DeepSeek-V3.2 | 80.8% | 46.0s | 282 | 2,152 |
| Qwen3.6 35B A3B | 80.7% | 25.4s | 340 | 5,144 |
| Qwen3.5-35B-A3B | 80.6% | 88.2s | 340 | 4,626 |
| GPT-5.4 mini | 80.5% | 12.3s | 290 | 2,194 |
| GPT-5.4 | 80.5% | 48.4s | 290 | 3,345 |
| Claude Sonnet 5 | 80.3% | 61.8s | 439 | 5,641 |
| GPT-5.6 Sol | 80.1% | 71.8s | 290 | 4,806 |
| GLM-5.2 | 79.5% | 50.1s | 295 | 2,524 |
| Qwen3.6 35B A3B | 79.4% | 20.1s | 340 | 4,805 |
| Qwen3.6 27B | 78.9% | 169.0s | 340 | 5,495 |
| Gemma 4 31B IT | 78.4% | 68.5s | 352 | 1,754 |
| Mistral Large 3 675B | 78.1% | 48.9s | 325 | 3,107 |
| Muse Glimmer 30B | 77.8% | 44.2s | 302 | 3,087 |
| Muse Glimmer 30B | 77.8% | 45.5s | 303 | 3,037 |
| MiniMax M2.7 | 77.8% | 61.3s | 287 | 4,984 |
| Qwen3.5-35B-A3B | 77.1% | 31.4s | 340 | 3,975 |
| DeepSeek-R1 | 77.1% | 150.8s | 283 | 2,956 |
| Qwen3.8 27B | 76.8% | 165.6s | 378 | 12,529 |
| Gemma 4 26B A4B IT | 76.6% | 12.0s | 338 | 1,537 |
| DeepSeek-V3.2 | 76.6% | 47.2s | 282 | 1,536 |
| GLM-5.3 | 76.1% | 24.5s | 296 | 2,818 |
| GPT-5.6 Sol | 75.8% | 63.1s | 290 | 4,292 |
| GLM-5.3 | 75.7% | 39.1s | 296 | 4,401 |
| GLM-5.2 | 75.4% | 38.4s | 295 | 2,606 |
| GLM-5.3-Flash | 75.1% | 449.5s | 295 | 14,294 |
| GPT-5.4 nano | 75.1% | 18.6s | 290 | 3,105 |
| Qwen3.6 35B A3B | 74.9% | 38.0s | 340 | 4,748 |
| GLM-5.2 | 74.3% | 89.8s | 295 | 3,105 |
| Qwen3.6 27B | 74.2% | 145.4s | 340 | 5,344 |
| DeepSeek-R1 | 74.1% | 166.7s | 283 | 3,300 |
| Qwen3.6 27B | 73.7% | 157.4s | 340 | 5,765 |
| Hy4 Preview | 73.4% | 241.8s | 303 | 16,357 |
| Qwen3.5-35B-A3B | 73.1% | 50.5s | 340 | 3,931 |
| DeepSeek-V3.2 | 73.1% | 53.2s | 282 | 1,421 |
| Claude Opus 5 | 73.0% | 277.5s | 439 | 20,669 |
| GPT-5.6 Terra | 72.7% | 53.3s | 290 | 3,569 |
| Qwen3.5-397B-A17B | 72.5% | 117.6s | 340 | 4,367 |
| Kimi K3 | 72.0% | 364.2s | 375 | 11,182 |
| DeepSeek V4 Pro | 72.0% | 95.5s | 282 | 4,137 |
| DeepSeek V4 Flash | 71.9% | 29.8s | 282 | 3,357 |
| Gemma 4 26B A4B IT | 71.9% | 28.8s | 338 | 1,479 |
| Gemma 4 26B A4B IT | 71.1% | 12.4s | 338 | 1,502 |
| Qwen3.8-Flash | 70.6% | 268.4s | 378 | 19,475 |
| Gemma 4 31B IT | 70.6% | 51.6s | 337 | 1,565 |
| DeepSeek V4 Pro | 70.6% | 56.9s | 282 | 3,378 |
| Mistral Small 2603 | 70.5% | 8.1s | 337 | 1,376 |
| Gemma 4 26B A4B IT | 70.3% | 20.2s | 337 | 1,500 |
| Gemma 4 31B IT | 69.8% | 51.6s | 337 | 1,586 |
| Gemma 4 31B IT | 69.3% | 61.8s | 337 | 1,692 |
| Gemini 3.1 Flash-Lite | 68.9% | 7.0s | 322 | 969 |
| Mistral Large 3 675B | 68.8% | 40.6s | 325 | 2,913 |
| DeepSeek V4 Pro | 68.1% | 93.0s | 282 | 5,096 |
| Qwen3.5-397B-A17B | 67.3% | 55.9s | 340 | 4,550 |
| Claude Opus 5 | 67.3% | 278.3s | 439 | 21,334 |
| Claude Opus 5 | 67.1% | 301.5s | 439 | 22,961 |
| Kimi K3 | 66.3% | 531.6s | 375 | 16,233 |
| Mistral Large 3 675B | 66.3% | 59.4s | 325 | 2,983 |
| MiniMax M3 | 66.0% | 33.9s | 392 | 4,392 |
| GPT-5.6 Luna | 65.8% | 40.1s | 290 | 3,592 |
| Llama 4 Maverick | 65.8% | 43.2s | 289 | 1,006 |
| DeepSeek-V3.2 | 65.6% | 39.6s | 282 | 1,810 |
| Kimi K3 | 65.0% | 345.8s | 375 | 11,058 |
| MiniMax M3 | 64.8% | 99.9s | 444 | 4,659 |
| Qwen3.5-397B-A17B | 64.6% | 46.9s | 340 | 4,157 |
| MiniMax M2.7 | 64.5% | 85.7s | 287 | 7,368 |
| MiniMax M2.7 | 64.1% | 103.0s | 287 | 8,687 |
| DeepSeek V4 Flash | 62.7% | 59.1s | 282 | 4,049 |
| DeepSeek-R1-Distill-Qwen-32B | 62.3% | 93.7s | 320 | 1,688 |
| Llama 3.3 70B Instruct | 62.3% | 7.8s | 295 | 881 |
| DeepSeek-R1 | 61.8% | 98.3s | 283 | 1,724 |
| Mistral Small 2603 | 61.5% | 21.7s | 337 | 2,245 |
| DeepSeek V4 Flash | 60.4% | 81.3s | 282 | 2,791 |
| Claude Haiku 4.5 | 60.3% | 72.2s | 332 | 9,054 |
| Claude Haiku 4.5 | 59.8% | 97.9s | 332 | 12,224 |
| Mistral Large 3 675B | 59.6% | 32.2s | 325 | 2,756 |
| Mistral Small 2603 | 57.8% | 16.4s | 337 | 1,724 |
| Llama 4 Maverick | 57.5% | 39.8s | 289 | 873 |
| Claude Haiku 4.5 | 57.5% | 76.0s | 332 | 9,670 |
| Llama 4 Scout | 57.0% | 11.2s | 289 | 855 |
| Llama 4 Scout | 56.3% | 26.4s | 289 | 818 |
| Llama 3.3 70B Instruct | 55.8% | 12.3s | 295 | 1,016 |
| Mistral Small 2603 | 54.5% | 16.2s | 337 | 2,044 |
| Llama 4 Scout | 51.7% | 25.1s | 289 | 873 |
| Llama 4 Maverick | 51.2% | 17.4s | 289 | 902 |
| Llama 4 Maverick | 51.0% | 42.2s | 289 | 930 |
| Llama 3.3 70B Instruct | 50.0% | 28.7s | 314 | 851 |
| Llama 4 Scout | 46.5% | 23.2s | 289 | 730 |
| Llama 3.3 70B Instruct | 44.8% | 12.1s | 295 | 967 |
| DeepSeek V4 Flash | 32.9% | 70.5s | 282 | 16,484 |
Tags
environmental_monitoringdeviationinvestigation