Gemma 4 E4B (Reasoning) vs Granite 4.2 3B

Which is better for the money: Gemma 4 E4B (Reasoning) by Google or Granite 4.2 3B by IBM? Intelligence, total benchmark cost, and end-to-end time from Artificial Analysis & other sources, snapshot 2026-10-24.

Trade-off

GoogleGemma 4 E4B (Reasoning)On the open-weights frontier

Intelligence: 8.90
Total cost: $14.7

IBMGranite 4.2 3BOn the open-weights frontier

Intelligence: 9.10
Total cost: $14.9

Verdict

Depends on your budget

Granite 4.2 3B scores 0.2 points higher (9.1 vs 8.9) but costs 1.0× as much to run the benchmark ($14.9 vs $14.7; both budget class). Pick Granite 4.2 3B when the extra intelligence is worth the premium, otherwise Gemma 4 E4B (Reasoning).

Metric	Gemma 4 E4B (Reasoning)	Granite 4.2 3B	Difference
Intelligence (higher is better)	8.9	9.1	+2.2% Granite 4.2 3B is 2.2% more intelligent
Total AA cost (lower is better)	$14.68	$14.91	+1.6% Granite 4.2 3B is 1.6% more expensive
End-to-end time (lower is better)	57.75s	35.30s	−38.9% Granite 4.2 3B is 38.9% faster
Price per 1M tokens (in / out)	$0.02 / $0.10	$0.03 / $0.12
Input modalities	Text, Image, Speech, Video	Text

Granite 4.2 3B vs Gemma 4 E4B (Reasoning): 2.2% more intelligence, 1.6% more cost to run the AA benchmark, and 38.9% faster end-to-end.

Open this matchup in the interactive picker Gemma 4 E4B (Reasoning): what beats it Granite 4.2 3B: what beats it

Today's Pareto frontier

Against the frontier

Gemma 4 E4B (Reasoning) is off the frontier: at the same or lower cost, GPT-6 Luna (Low) scores 21.50 for $10.6.
Granite 4.2 3B is off the frontier: at the same or lower cost, GPT-6 Luna (Low) scores 21.50 for $10.6.

Gemma 4 E4B (Reasoning)Granite 4.2 3BToday's Pareto frontierOpen-weights frontierEvery other model

Matchup

Compare something else

Agents: Accept: text/markdown or ?format=md returns this page as Markdown. JSON: /api/v1/compare/gemma-4-e4b/granite-4-2-3b · Markdown: /api/v1/compare/gemma-4-e4b/granite-4-2-3b?format=md · llms.txt · agent skill