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

Kimi Linear 48B A3B Instruct

  • Kimi
  • Release · October 30, 2025
  • Open weights
  • Non-reasoning

Intelligence

Intelligence Index

#209 of 266
  1. 7.57.5Hermes 4 - Llama-3.1 405B (Reasoning)
  2. 7.47.4Hermes 4 - Llama-3.1 405B (Non-reasoning)
  3. 7.47.4NVIDIA Nemotron Nano 9B V2 (Reasoning)
  4. 7.37.3K2-V2 (Low)
  5. 7.37.3Llama 3.1 Instruct 405B
  6. 7.37.3Kimi Linear 48B A3B Instruct
  7. 7.27.2LFM2.5-8B-A1B
  8. 7.27.2Ring-flash-2.0
  9. 7.17.1Olmo 3.1 32B Think
  10. 77Command A
Data from Artificial Analysis

Evaluations

  • Humanity's Last Exam2.5%#263
  • AA-LCR28%#177
  • GPQA Diamond41.2%#184
  • IFBench28.1%#131
  • τ²-Bench Telecom0%#143
  • Terminal-Bench Hard11.4%#53
Kimi Linear 48B A3B InstructMedian
Data from Artificial Analysis

Nearby models

  • Hermes 4 - Llama-3.1 405B (Reasoning)Nous Research7.5
  • Hermes 4 - Llama-3.1 405B (Non-reasoning)Nous Research7.4
  • NVIDIA Nemotron Nano 9B V2 (Reasoning)NVIDIA7.4
  • K2-V2 (Low)Institute of Foundation Models7.3
  • Llama 3.1 Instruct 405BMeta7.3
  • LFM2.5-8B-A1BLiquid AI7.2
  • Ring-flash-2.0InclusionAI7.2
  • Olmo 3.1 32B ThinkAllen Institute for AI7.1
Data from Artificial Analysis

Same lab

  • Kimi K3.1 ProKimi55.1
  • Kimi K2.7 CodeKimi25.8
Data from Artificial Analysis
HorizonVascend IDEOlympusTalosCoursesCommunityBenchmark
DiscordGitHubXLet’s talkLegalPrivacyCookiesData useSecurity
© 2026 Vascend Technologies · VAT IT04911550616
ascend