Gemma 4 E4B

Composite indexes

• Artificial Analysis Intelligence Index#106 of 142
  9independentv4.3.2, reasoningArtificial Analysis: Gemma 4 E4B (Reasoning) ↗
  8independentv4.3.2, non-reasoningArtificial Analysis: Gemma 4 E4B (Non-reasoning) ↗

  Leader: Claude Opus 5.5 at 58

Coding

• Terminal-Bench 2.x#100 of 103
  1.9%independentAA run (Terminal-Bench 2.1), reasoningArtificial Analysis: Gemma 4 E4B (Reasoning) ↗

  Leader: Claude Fable 5.1 at 91.4%

• LiveCodeBench#33 of 56
  52%vendorLiveCodeBench v6Gemma 4 model card ↗

  Leader: Fugu at 92.9%

Agents and tool use

• GDPval-AA#98 of 103
  -51independentGDPval-AA v2.1, reasoningArtificial Analysis: Gemma 4 E4B (Reasoning) ↗

  Leader: Claude Opus 5.5 at 1846

• τ²-Bench#66 of 88
  42.2%vendoraverage over 3 domainsGemma 4 model card ↗
  20.8%independentAA run, telecom, reasoningArtificial Analysis: Gemma 4 E4B (Reasoning) ↗

  Leader: GLM-5.2 at 99.1%

Reasoning and knowledge

• GPQA Diamond#114 of 142
  58.6%vendorGemma 4 model card ↗
  57.6%independentAA run, reasoningArtificial Analysis: Gemma 4 E4B (Reasoning) ↗

  Leader: GPT-6 Astra at 96.1%

• Humanity’s Last Exam#141 of 147
  3.8%independentAA run, no tools, reasoningArtificial Analysis: Gemma 4 E4B (Reasoning) ↗

  Leader: Claude Opus 5.5 at 61.4%

• CritPt#74 of 139
  0.6%independentAA run, reasoningArtificial Analysis: Gemma 4 E4B (Reasoning) ↗

  Leader: GPT-5.6 Sol at 32.3%

• MMLU-Pro#12 of 14
  69.4%vendorinstruction-tunedGemma 4 model card ↗

  Leader: Nemotron 3 Ultra 550B A55B at 86.8%

• AIME 2026#12 of 13
  42.5%vendorno toolsGemma 4 model card ↗

  Leader: ERNIE 5.1 at 99.6%

Vision and long context

• AA Long Context Reasoning#105 of 139
  32%independentAA-LCR v1.1, reasoningArtificial Analysis: Gemma 4 E4B (Reasoning) ↗

  Leader: Kimi K3 at 88.7%

• MMMU-Pro#64 of 72
  52.6%vendorGemma 4 model card ↗
  51.4%independentAA run, reasoningArtificial Analysis: Gemma 4 E4B (Reasoning) ↗

  Leader: Claude Opus 5.5 at 87.7%

Other published results