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

Gemini 3.1 Pro Preview

  • Google
  • Release · February 19, 2026
  • Proprietary
  • Reasoning

Intelligence

Intelligence Index

#72 of 266
  1. 32.533GPT-5.3 Codex (Xhigh)
  2. 30.531K2 Horizon 375B A23B
  3. 30.330Beam 1.0
  4. 30.130GPT-5.6 Terra (Medium)
  5. 29.930GPT-6 Luna (Medium)
  6. 29.730Gemini 3.1 Pro Preview
  7. 29.530Claude Haiku 5.5 (Low)
  8. 29.229MiniMax-M3
  9. 28.228Nex-N2-Pro (Based on Qwen3.5-397B-A17B)
  10. 27.628Qwen3.8 27B (Medium)
Data from Artificial Analysis

Evaluations

  • AA-Omniscience Accuracy54.9%#22
  • Humanity's Last Exam47%#30
  • Terminal-Bench 4.04%#66
  • AA-LCR82%#31
  • SciCode58.7%#15
  • CritPt17.7%#45
  • AA-AnalystAgent41.3%#13
  • MMMU-Pro82.4%#21
  • GPQA Diamond94.1%#5
  • IFBench77.1%#8
  • τ²-Bench Telecom95.6%#2
  • τ²-Bench Banking21.4%#43
  • Terminal-Bench Hard53.8%#4
  • APEX-Agents32%#4
Gemini 3.1 Pro PreviewMedian
Data from Artificial Analysis

Speed and price

Speed

tokens per second · #54 of 162
  1. 140140GPT-6 Luna (Max)
  2. 137137Gemini 3.8 Flash (High)
  3. 136136Apertus 8B Instruct
  4. 134134Claude Sonnet 5.5 (Max)
  5. 133133Qwen3 Next 80B A3B Instruct
  6. 132132Gemini 3.1 Pro Preview
  7. 131131Qwen3.6 35B A3B (Reasoning)
  8. 130130GPT-5.6 Terra (Max)
  9. 128128Qwen3.5 122B A10B (Reasoning)
  10. 127127GPT-6 Luna (Xhigh)
Data from Artificial Analysis

Latency

seconds to first token · #138 of 162
  1. 15.115GPT-5.6 Terra (Xhigh)
  2. 15.716Claude Opus 5.5 (Medium)
  3. 16.617Nova 2.0 Lite (High)
  4. 19.820GPT-6 Luna (Xhigh)
  5. 2020Claude Fable 5.1 (High)
  6. 20.621Gemini 3.1 Pro Preview
  7. 20.921Grok 4.7 (High)
  8. 2525Claude Haiku 5.5 (High)
  9. 27.327Gemini 3.8 Flash (High)
  10. 27.327Claude Sonnet 5.5 (Xhigh)
Data from Artificial Analysis

Price

$ per 1M tokens (3:1) · #149 of 169
  1. $4.384.38Command A
  2. $4.54.5GPT-5.6 Terra (Max)
  3. $4.54.5GPT-5.6 Terra (Xhigh)
  4. $4.54.5GPT-5.6 Terra (High)
  5. $4.54.5GPT-5.6 Terra (Medium)
  6. $4.54.5Gemini 3.1 Pro Preview
  7. $4.54.5GPT-5.6 Terra (Low)
  8. $4.54.5GPT-5.6 Terra (Non-reasoning)
  9. $4.814.81GPT-5.3 Codex (Xhigh)
  10. $88Claude Opus 5.5 (Max)
Data from Artificial Analysis

Nearby models

  • GPT-5.3 Codex (Xhigh)OpenAI32.5
  • K2 Horizon 375B A23BInstitute of Foundation Models30.5
  • Beam 1.0Reflection30.3
  • GPT-5.6 Terra (Medium)OpenAI30.1
  • GPT-6 Luna (Medium)OpenAI29.9
  • Claude Haiku 5.5 (Low)Anthropic29.5
  • MiniMax-M3MiniMax29.2
  • Nex-N2-Pro (Based on Qwen3.5-397B-A17B)Nex AGI28.2
Data from Artificial Analysis

Same lab

  • Gemini 4 Argon (High)Google52.6
  • Gemini 3.8 Flash (High)Google40.9
  • Gemini 3.8 Flash (Medium)Google39.8
  • Gemini 3.8 Flash (Low)Google33.5
  • Gemini 3.5 Flash-LiteGoogle22.2
  • Gemma 4 26B A4B (Reasoning)Google16.7
  • Gemma 4 31B (Reasoning)Google14.7
  • Gemma 4 12B (Reasoning)Google14.2
Data from Artificial Analysis
HorizonVascend IDEOlympusTalosCoursesCommunityBenchmark
DiscordGitHubXLet’s talkLegalPrivacyCookiesData useSecurity
© 2026 Vascend Technologies · VAT IT04911550616
ascend