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

Phi-4 Mini Instruct

  • Microsoft
  • Release · February 26, 2024
  • Open weights
  • Non-reasoning

Intelligence

Intelligence Index

#226 of 266
  1. 6.56.5Olmo 3.1 32B Instruct
  2. 6.46.4R1 1776
  3. 6.46.4Llama 3.2 Instruct 90B (Vision)
  4. 6.36.3EXAONE 4.0 32B (Non-reasoning)
  5. 6.36.3Celeris-1
  6. 6.36.3Phi-4 Mini Instruct
  7. 6.26.2Qwen3.5 2B (Non-reasoning)
  8. 6.16.1DeepHermes 3 - Mistral 24B Preview (Non-reasoning)
  9. 6.16.1Jamba 1.7 Large
  10. 6.16.1Qwen3.5 0.8B (Reasoning)
Data from Artificial Analysis

Evaluations

  • AA-Omniscience Accuracy9.5%#224
  • Humanity's Last Exam4.5%#229
  • AA-LCR15.3%#206
  • CritPt0%#143
  • GPQA Diamond33.1%#195
  • IFBench21.1%#148
  • τ²-Bench Telecom8.2%#141
  • Terminal-Bench Hard0%#121
Phi-4 Mini InstructMedian
Data from Artificial Analysis

Speed and price

Speed

tokens per second · #149 of 162
  1. 5151GPT-6.1 Sol (Medium)
  2. 5050Qwen3.8 27B (Non-reasoning)
  3. 4949Qwen3.8 27B (Low)
  4. 4848Qwen3.8 27B (Medium)
  5. 4848Granite 4.2 30B
  6. 4646Phi-4 Mini Instruct
  7. 4545Gemma 4 E4B (Reasoning)
  8. 4545Qwen3.8 27B (Xhigh)
  9. 4545Gemma 4 E4B (Non-reasoning)
  10. 4343Hermes 4 - Llama-3.1 405B (Non-reasoning)
Data from Artificial Analysis

Latency

seconds to first token · #23 of 162
  1. 0.80.8Gemma 4 E4B (Non-reasoning)
  2. 0.80.8NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning)
  3. 0.80.8Nova Micro
  4. 0.80.8Granite 4.2 30B
  5. 0.80.8Nemotron 3 Nano Omni 30B A3B Reasoning
  6. 0.90.9Phi-4 Mini Instruct
  7. 0.90.9Ministral 3 8B
  8. 0.90.9Mistral Small 4 (Reasoning)
  9. 0.90.9GPT-6 Luna (Non-reasoning)
  10. 0.90.9GPT-5.5 Instant (June 2026)
Data from Artificial Analysis

Price

$ per 1M tokens (3:1) · #24 of 169
  1. $0.10.1Nemotron 3.5 Lightning
  2. $0.10.1Ministral 3 3B
  3. $0.110.11Granite 4.2 8B
  4. $0.130.13Apertus 8B Instruct
  5. $0.130.13gpt-oss-20b (Low)
  6. $0.130.13Phi-4 Mini Instruct
  7. $0.140.14Phi-4 Multimodal Instruct
  8. $0.150.15Agnes 2.5 Pro Beta
  9. $0.150.15Gemma 4 12B (Reasoning)
  10. $0.150.15Gemma 4 12B (Non-reasoning)
Data from Artificial Analysis

Nearby models

  • Olmo 3.1 32B InstructAllen Institute for AI6.5
  • R1 1776Perplexity6.4
  • Llama 3.2 Instruct 90B (Vision)Meta6.4
  • EXAONE 4.0 32B (Non-reasoning)LG AI Research6.3
  • Celeris-1Celeris6.3
  • Qwen3.5 2B (Non-reasoning)Alibaba6.2
  • DeepHermes 3 - Mistral 24B Preview (Non-reasoning)Nous Research6.1
  • Jamba 1.7 LargeAI21 Labs6.1
Data from Artificial Analysis

Same lab

  • Phi-4Microsoft5.9
  • Phi-4 Multimodal InstructMicrosoft5.8
Data from Artificial Analysis
HorizonVascend IDEOlympusTalosCoursesCommunityBenchmark
DiscordGitHubXLet’s talkLegalPrivacyCookiesData useSecurity
© 2026 Vascend Technologies · VAT IT04911550616
ascend