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

Qwen3.8 2.4T A95B

  • Alibaba
  • Release · August 12, 2026
  • Open weights
  • Reasoning

Intelligence

Intelligence Index

#45 of 266
  1. 41.842GLM-5.3-Flash
  2. 41.241Claude Haiku 5.5 (Xhigh)
  3. 41.141Ling 3.1 Flash
  4. 40.941Gemini 3.8 Flash (High)
  5. 40.841Claude Sonnet 5.5 (Medium)
  6. 39.940Qwen3.8 2.4T A95B
  7. 39.840Gemini 3.8 Flash (Medium)
  8. 39.840Qwen3.8-Flash-Next
  9. 38.939Muse Spark Open (Xhigh)
  10. 38.438Mistral Large 4 Preview
Data from Artificial Analysis

Evaluations

  • AA-Omniscience Accuracy31.3%#76
  • Humanity's Last Exam42.5%#42
  • Terminal-Bench 4.011.1%#58
  • AA-LCR80.3%#41
  • SciCode54.1%#41
  • CritPt20%#36
  • GPQA Diamond93.5%#8
  • τ²-Bench Banking49.1%#2
Qwen3.8 2.4T A95BMedian
Data from Artificial Analysis

Speed and price

Speed

tokens per second · #155 of 162
  1. 4545Gemma 4 E4B (Reasoning)
  2. 4545Qwen3.8 27B (Xhigh)
  3. 4545Gemma 4 E4B (Non-reasoning)
  4. 4343Hermes 4 - Llama-3.1 405B (Non-reasoning)
  5. 4343Hermes 4 - Llama-3.1 405B (Reasoning)
  6. 4141Qwen3.8 2.4T A95B
  7. 4040Qwen3.8 Max (0902)
  8. 3939Phi-4
  9. 3535Gemma 4 31B (Reasoning)
  10. 3333Command A
Data from Artificial Analysis

Latency

seconds to first token · #106 of 162
  1. 2.82.8Hy3
  2. 2.82.8GPT-6 Luna (Medium)
  3. 2.82.8Kimi K2.7 Code
  4. 2.82.8Qwen3.8 Max (0902)
  5. 2.82.8GLM-5.3-Flash
  6. 2.92.9Qwen3.8 2.4T A95B
  7. 33GLM-5.3 (Low)
  8. 3.13.1Kimi K3.1 Pro
  9. 3.13.1MiMo-V2.6-Pro
  10. 3.53.5Apertus 8B Instruct
Data from Artificial Analysis

Price

$ per 1M tokens (3:1) · #126 of 169
  1. $2.152.15GLM-5.3 (Low)
  2. $33Grok 4.7 (Xhigh)
  3. $33Grok 4.7 (High)
  4. $33Qwen3.8 Max (0902)
  5. $33Grok 4.7 (Low)
  6. $33Qwen3.8 2.4T A95B
  7. $33Mistral Medium 3.5
  8. $3.443.44Nova 2.0 Pro Preview (Medium)
  9. $3.443.44Nova 2.0 Pro Preview (Low)
  10. $3.443.44Nova 2.0 Pro Preview (Non-reasoning)
Data from Artificial Analysis

Nearby models

  • GLM-5.3-FlashZ AI41.8
  • Claude Haiku 5.5 (Xhigh)Anthropic41.2
  • Ling 3.1 FlashInclusionAI41.1
  • Gemini 3.8 Flash (High)Google40.9
  • Claude Sonnet 5.5 (Medium)Anthropic40.8
  • Gemini 3.8 Flash (Medium)Google39.8
  • Qwen3.8-Flash-NextAlibaba39.8
  • Muse Spark Open (Xhigh)Meta38.9
Data from Artificial Analysis

Same lab

  • Qwen3.8 Max (0902)Alibaba45.4
  • Qwen3.8 27B (Xhigh)Alibaba33.7
  • Qwen3.8 27B (Medium)Alibaba27.6
  • Qwen3.8 27B (Low)Alibaba26.2
  • Qwen3.7 PlusAlibaba25.2
  • Qwen3.5 397B A17B (Non-reasoning)Alibaba21.4
  • Qwen3.5 Omni PlusAlibaba20.4
  • Qwen3.8 27B (Non-reasoning)Alibaba20.2
Data from Artificial Analysis
HorizonVascend IDEOlympusTalosCoursesCommunityBenchmark
DiscordGitHubXLet’s talkLegalPrivacyCookiesData useSecurity
© 2026 Vascend Technologies · VAT IT04911550616
ascend