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

GLM-5.3-Flash

  • Z AI
  • Release · August 26, 2026
  • Open weights
  • Reasoning

Intelligence

Intelligence Index

#40 of 266
  1. 42.743MiniMax-M3.1-Flash
  2. 42.342Claude Opus 5.5 (Low)
  3. 42.242Grok 4.7 (Low)
  4. 42.142GPT-5.6 Terra (Max)
  5. 42.142GPT-6.1 Sol (Low)
  6. 41.842GLM-5.3-Flash
  7. 41.241Claude Haiku 5.5 (Xhigh)
  8. 41.141Ling 3.1 Flash
  9. 40.941Gemini 3.8 Flash (High)
  10. 40.841Claude Sonnet 5.5 (Medium)
Data from Artificial Analysis

Evaluations

  • AA-Briefcase1,452#17
  • AA-Omniscience Accuracy27.5%#84
  • AA-Omniscience Non-Hallucination72.4%#6
  • AutomationBench-AA60.4%#13
  • GDPval-AA1,642#12
  • Humanity's Last Exam39.9%#49
  • GDP.pdf15.4%#21
  • Terminal-Bench 4.032.8%#34
  • AA-LCR80%#44
  • SciCode51.6%#56
  • CritPt15.4%#54
  • Terminal-Bench-Science4.8%#17
  • FrontierCode41%#16
  • AA-AnalystAgent25%#19
  • Harvey LAB-AA0.8%#12
  • EnterpriseOps-Gym-AA33.2%#6
  • AA-3D937#14
  • ITBench-AA51.2%#3
  • MLCR-AA11.7%#21
  • PersonalityBench: Expressiveness95#3
  • PersonalityBench: Organization35.6#21
  • PersonalityBench: Transparency70.3#10
  • PersonalityBench: Slop33.2#15
  • PersonalityBench: Informativeness60.4#14
  • PersonalityBench: Warmth85.1#5
  • PersonalityBench: Conviction80.2#6
  • PersonalityBench: Imagination77.7#7
  • GPQA Diamond91.2%#21
  • τ²-Bench Banking47.2%#6
GLM-5.3-FlashMedian
Data from Artificial Analysis

Cost per task

Cost per task

$ per Intelligence Index task · #8 of 24
  1. $0.190.19Nemotron 3.5 Super VL
  2. $0.210.21Claude Haiku 5.5 (Max)
  3. $0.220.22MiniMax-M3.1-Flash
  4. $0.230.23Beam 1.0
  5. $0.260.26DeepSeek V4.1 Flash 1017 (Max)
  6. $0.280.28GLM-5.3-Flash
  7. $0.720.72GPT-6.1 Sol (Max)
  8. $0.840.84Muse Spark Open (Xhigh)
  9. $1.011.01Qwen3.8 27B (Xhigh)
  10. $1.031.03Step 5 Preview
Data from Artificial Analysis

Time per task

seconds per task · #20 of 26
  1. 604604GLM-5.3 (Max)
  2. 628628GPT-6.1 Sol (Max)
  3. 723723Gemini 4 Argon (High)
  4. 736736Claude Fable 5.1 (Max)
  5. 791791Claude Opus 5.5 (Max)
  6. 824824GLM-5.3-Flash
  7. 846846Kimi K3.1 Pro
  8. 901901Grok 4.7 (Xhigh)
  9. 936936Claude Sonnet 5.5 (Max)
  10. 1,0131,013MiMo-V2.6-Pro
Data from Artificial Analysis

Tokens per task

output tokens per task · #16 of 27
  1. 55.2K55KKimi K3.1 Pro
  2. 61.6K62KGemini 4 Argon (High)
  3. 64.1K64KStep 5 Preview
  4. 64.3K64KMiMo-V2.6-Pro
  5. 66.8K67KQwen3.8 27B (Xhigh)
  6. 68.7K69KGLM-5.3-Flash
  7. 69.5K70KMistral Large 4 Preview
  8. 71K71KGemini 3.8 Flash (High)
  9. 71.1K71KGLM-5.3 (Max)
  10. 78.1K78KClaude Fable 5.1 (Max)
Data from Artificial Analysis

Speed and price

Speed

tokens per second · #114 of 162
  1. 8080Qwen3.5 9B (Non-reasoning)
  2. 7878Grok 4.7 (Low)
  3. 7777Ministral 3 14B
  4. 7777Magistral Small 1.2
  5. 7676LongCat 2.0
  6. 7070GLM-5.3-Flash
  7. 7070Claude Fable 5.1 (Max)
  8. 6666GPT-6 Astra (Max)
  9. 6666North Mini Code
  10. 6464Kimi K2.7 Code
Data from Artificial Analysis

Latency

seconds to first token · #105 of 162
  1. 2.62.6Ling 3.0 Tiny
  2. 2.82.8Hy3
  3. 2.82.8GPT-6 Luna (Medium)
  4. 2.82.8Kimi K2.7 Code
  5. 2.82.8Qwen3.8 Max (0902)
  6. 2.82.8GLM-5.3-Flash
  7. 2.92.9Qwen3.8 2.4T A95B
  8. 33GLM-5.3 (Low)
  9. 3.13.1Kimi K3.1 Pro
  10. 3.13.1MiMo-V2.6-Pro
Data from Artificial Analysis

Price

$ per 1M tokens (3:1) · #52 of 169
  1. $0.20.2GPT-6 Luna (Non-reasoning)
  2. $0.20.2Ministral 3 14B
  3. $0.220.22Phi-4
  4. $0.220.22Gemma 4 31B (Non-reasoning)
  5. $0.230.23Qwen3.8-Flash-Next
  6. $0.240.24GLM-5.3-Flash
  7. $0.240.24Hy3
  8. $0.240.24gpt-oss-120b (Low)
  9. $0.260.26MiniMax-M3.1-Flash
  10. $0.260.26gpt-oss-120b (High)
Data from Artificial Analysis

Nearby models

  • MiniMax-M3.1-FlashMiniMax42.7
  • Claude Opus 5.5 (Low)Anthropic42.3
  • Grok 4.7 (Low)SpaceXAI42.2
  • GPT-5.6 Terra (Max)OpenAI42.1
  • GPT-6.1 Sol (Low)OpenAI42.1
  • Claude Haiku 5.5 (Xhigh)Anthropic41.2
  • Ling 3.1 FlashInclusionAI41.1
  • Gemini 3.8 Flash (High)Google40.9
Data from Artificial Analysis

Same lab

  • GLM-5.3 (Max)Z AI44.8
  • GLM-5.3 (Low)Z AI34.3
Data from Artificial Analysis
HorizonVascend IDEOlympusTalosCoursesCommunityBenchmark
DiscordGitHubXLet’s talkLegalPrivacyCookiesData useSecurity
© 2026 Vascend Technologies · VAT IT04911550616
ascend