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

Granite 4.0 H 350M

  • IBM
  • Release · October 28, 2025
  • Open weights
  • Non-reasoning

Intelligence

Intelligence Index

#259 of 266
  1. 5.15.1Gemma 3 270M
  2. 5.15.1DeepHermes 3 - Llama-3.1 8B Preview (Non-reasoning)
  3. 5.15.1Apertus 70B Instruct
  4. 5.15.1Granite 4.0 Micro
  5. 55Molmo2-8B
  6. 4.84.8Granite 4.0 H 350M
  7. 4.84.8Granite 4.0 350M
  8. 4.84.8NVIDIA Nemotron 3 Nano 4B
  9. 4.84.8Ministral 3 3B
  10. 4.84.8Tiny Aya Global
Data from Artificial Analysis

Evaluations

  • AA-Omniscience Accuracy3.8%#256
  • Humanity's Last Exam6.4%#181
  • AA-LCR0%#229
  • CritPt0%#143
  • GPQA Diamond25.7%#209
  • IFBench17.6%#151
  • τ²-Bench Telecom14.6%#126
  • Terminal-Bench Hard0%#121
Granite 4.0 H 350MMedian
Data from Artificial Analysis

Nearby models

  • Gemma 3 270MGoogle5.1
  • DeepHermes 3 - Llama-3.1 8B Preview (Non-reasoning)Nous Research5.1
  • Apertus 70B InstructSwiss AI Initiative5.1
  • Granite 4.0 MicroIBM5.1
  • Molmo2-8BAllen Institute for AI5
  • Granite 4.0 350MIBM4.8
  • NVIDIA Nemotron 3 Nano 4BNVIDIA4.8
  • Ministral 3 3BMistral4.8
Data from Artificial Analysis

Same lab

  • Granite 4.2 30BIBM12.7
  • Granite 4.2 8BIBM11.1
  • Granite 4.2 3BIBM9.1
  • Granite 4.0 H SmallIBM6
  • Granite 4.0 H 1BIBM5.2
Data from Artificial Analysis
HorizonVascend IDEOlympusTalosCoursesCommunityBenchmark
DiscordGitHubXLet’s talkLegalPrivacyCookiesData useSecurity
© 2026 Vascend Technologies · VAT IT04911550616
ascend