BenchmarkInkling (Xhigh)
- Thinking Machines
- Rilascio · 15 luglio 2026
- Pesi aperti
- Ragionamento
Intelligenza
- AA-Briefcase832#27
- AA-Omniscience Accuracy41,6%#51
- AA-Omniscience Non-Hallucination32,3%#24
- AutomationBench-AA5%#27
- GDPval-AA1072#27
- Humanity's Last Exam31,9%#80
- GDP.pdf12,8%#23
- Terminal-Bench 4.01%#77
- AA-LCR77,3%#73
- SciCode47%#80
- CritPt5,4%#81
- Terminal-Bench-Science0%#25
- FrontierCode15,8%#27
- AA-AnalystAgent23,8%#20
- Harvey LAB-AA0%#15
- EnterpriseOps-Gym-AA38%#4
- AA-3D583#18
- MLCR-AA6,1%#25
- MMMU-Pro73,5%#55
- PersonalityBench: Expressiveness77,7#9
- PersonalityBench: Organization92,6#3
- PersonalityBench: Transparency53#15
- PersonalityBench: Slop50,5#12
- PersonalityBench: Informativeness62,9#13
- PersonalityBench: Warmth45,6#16
- PersonalityBench: Conviction67,8#9
- PersonalityBench: Imagination67,8#10
- GPQA Diamond87,2%#38
- τ²-Bench Banking29,1%#32
Inkling (Xhigh)Mediana
Dati di Artificial AnalysisCosto per task
Velocità e prezzo
Latenza
secondi al primo token · #69 su 162- 1,71,7
Claude Fable 5.1 (Low) - 1,81,8
DeepSeek V4 Pro 0813 (Max) - 1,81,8
Llama 3.3 Instruct 70B - 1,81,8
DeepSeek V4 Pro 0813 (Non-reasoning) - 1,81,8
Mistral Large 4 Preview - 1,81,8Inkling (Xhigh)
- 1,81,8
Qwen3.5 Omni Flash - 1,81,8
Command A - 1,81,8
Ling 3.1 Flash - 1,81,8
Gemma 4 31B (Non-reasoning)
Dati di Artificial AnalysisPrezzo
$ per 1M token (3:1) · #115 su 169- $1,51,5
Gemini 3.8 Flash (Low) - $1,51,5
Qwen3.5 Omni Plus - $1,51,5
Hermes 4 - Llama-3.1 405B (Reasoning) - $1,51,5
Hermes 4 - Llama-3.1 405B (Non-reasoning) - $1,711,71
Kimi K2.7 Code - $1,761,76Inkling (Xhigh)
- $1,981,98
DeepSeek V4 Pro 0813 (Max) - $1,981,98
DeepSeek V4 Pro 0813 (Non-reasoning) - $22
Muse Spark Open (Xhigh) - $2,072,07
Mistral Large 4 Preview
Dati di Artificial AnalysisModelli vicini