Vascend
HorizonVascend IDEVascend CLICommunityBenchmarkContact
Sign inJoin Discord
HomeHorizonIDECLISign in
ChangelogIntelligenceCost per taskCoding Agent IndexCyber IndexDecision IndexBenchmarksSpeedLatencyPriceLeaderboard
Benchmark

JT-4.1 Flash 236B A21B (Reasoning)

  • China Mobile
  • Release · July 9, 2026
  • Proprietary
  • Reasoning

Intelligence

Intelligence Index

#62 of 266
  1. 34.835DeepSeek V4 Flash Vision (Max)
  2. 34.635GPT-6 Luna (Xhigh)
  3. 34.535Claude Haiku 5.5 (Medium)
  4. 34.334GLM-5.3 (Low)
  5. 34.234GPT-5.6 Terra (High)
  6. 33.934JT-4.1 Flash 236B A21B (Reasoning)
  7. 33.734Qwen3.8 27B (Xhigh)
  8. 33.634Motif 3
  9. 33.534Gemini 3.8 Flash (Low)
  10. 32.933GPT-6 Luna (High)
Data from Artificial Analysis

Evaluations

  • AA-Omniscience Accuracy49.4%#33
  • Humanity's Last Exam43.6%#35
  • Terminal-Bench 4.03%#67
  • AA-LCR78.3%#65
  • SciCode52%#53
  • CritPt14%#58
JT-4.1 Flash 236B A21B (Reasoning)Median
Data from Artificial Analysis

Nearby models

  • DeepSeek V4 Flash Vision (Max)DeepSeek34.8
  • GPT-6 Luna (Xhigh)OpenAI34.6
  • Claude Haiku 5.5 (Medium)Anthropic34.5
  • GLM-5.3 (Low)Z AI34.3
  • GPT-5.6 Terra (High)OpenAI34.2
  • Qwen3.8 27B (Xhigh)Alibaba33.7
  • Motif 3Motif Technologies33.6
  • Gemini 3.8 Flash (Low)Google33.5
Data from Artificial Analysis

Same lab

  • JT-4.1 Flash 236B A21B (Non-reasoning)China Mobile27.3
  • JT-35B-FlashChina Mobile18.7
  • JT-MINIChina Mobile12.2
Data from Artificial Analysis
HorizonVascend IDEOlympusTalosCoursesCommunityBenchmark
DiscordGitHubXLet’s talkLegalPrivacyCookiesData useSecurity
© 2026 Vascend Technologies · VAT IT04911550616
ascend