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

Hy3

  • Tencent
  • Release · July 6, 2026
  • Open weights
  • Reasoning

Intelligence

Intelligence Index

#87 of 266
  1. 26.426Apodex 1.1
  2. 26.226Qwen3.8 27B (Low)
  3. 2626GPT-5.5 Instant (June 2026)
  4. 25.826Kimi K2.7 Code
  5. 25.726Inkling Small
  6. 25.325Hy3
  7. 25.325K2 Horizon MoVA 36B A4B
  8. 25.225Qwen3.7 Plus
  9. 2525Inkling (Xhigh)
  10. 24.725Solar Open2 250B
Data from Artificial Analysis

Evaluations

  • AA-Omniscience Accuracy32%#72
  • Humanity's Last Exam33.5%#75
  • Terminal-Bench 4.00.5%#86
  • AA-LCR79%#63
  • SciCode48.6%#71
  • CritPt4.9%#83
  • GPQA Diamond89.7%#27
  • τ²-Bench Banking22.9%#40
Hy3Median
Data from Artificial Analysis

Speed and price

Speed

tokens per second · #107 of 162
  1. 8686Grok 4.7 (High)
  2. 8383Claude Opus 5.5 (Medium)
  3. 8282Claude Opus 5.5 (Xhigh)
  4. 8282Claude Opus 5.5 (Low)
  5. 8181Gemini 4 Argon (High)
  6. 8181Hy3
  7. 8080Claude Opus 5.5 (High)
  8. 8080Qwen3.5 9B (Non-reasoning)
  9. 7878Grok 4.7 (Low)
  10. 7777Ministral 3 14B
Data from Artificial Analysis

Latency

seconds to first token · #101 of 162
  1. 2.42.4GLM-5.3 (Max)
  2. 2.52.5GPT-6 Luna (Low)
  3. 2.52.5Phi-4
  4. 2.62.6LongCat 2.0
  5. 2.62.6Ling 3.0 Tiny
  6. 2.82.8Hy3
  7. 2.82.8GPT-6 Luna (Medium)
  8. 2.82.8Kimi K2.7 Code
  9. 2.82.8Qwen3.8 Max (0902)
  10. 2.82.8GLM-5.3-Flash
Data from Artificial Analysis

Price

$ per 1M tokens (3:1) · #53 of 169
  1. $0.20.2Ministral 3 14B
  2. $0.220.22Phi-4
  3. $0.220.22Gemma 4 31B (Non-reasoning)
  4. $0.230.23Qwen3.8-Flash-Next
  5. $0.240.24GLM-5.3-Flash
  6. $0.240.24Hy3
  7. $0.240.24gpt-oss-120b (Low)
  8. $0.260.26MiniMax-M3.1-Flash
  9. $0.260.26gpt-oss-120b (High)
  10. $0.260.26Mistral Small 4 (Reasoning)
Data from Artificial Analysis

Nearby models

  • Apodex 1.1Apodex26.4
  • Qwen3.8 27B (Low)Alibaba26.2
  • GPT-5.5 Instant (June 2026)OpenAI26
  • Kimi K2.7 CodeKimi25.8
  • Inkling SmallThinking Machines25.7
  • K2 Horizon MoVA 36B A4BInstitute of Foundation Models25.3
  • Qwen3.7 PlusAlibaba25.2
  • Inkling (Xhigh)Thinking Machines25
Data from Artificial Analysis
HorizonVascend IDEOlympusTalosCoursesCommunityBenchmark
DiscordGitHubXLet’s talkLegalPrivacyCookiesData useSecurity
© 2026 Vascend Technologies · VAT IT04911550616
ascend