Granite 4.2 3B vs Llama 4 Maverick

Which is better for the money: Granite 4.2 3B by IBM or Llama 4 Maverick by Meta? Intelligence, total benchmark cost, and end-to-end time from Artificial Analysis & other sources, snapshot 2026-10-25.

Trade-off

IBMGranite 4.2 3BOn the open-weights frontier

Intelligence: 9.10
Total cost: $14.9

MetaLlama 4 MaverickOff the frontier

Intelligence: 9.30
Total cost: $44.0

Verdict

Depends on your budget

Llama 4 Maverick scores 0.2 points higher (9.3 vs 9.1) but costs 2.9× as much to run the benchmark ($44.0 vs $14.9; both budget class). Pick Llama 4 Maverick when the extra intelligence is worth the premium, otherwise Granite 4.2 3B.

Metric	Granite 4.2 3B	Llama 4 Maverick	Difference
Intelligence (higher is better)	9.1	9.3	+2.2% Llama 4 Maverick is 2.2% more intelligent
Total AA cost (lower is better)	$14.91	$43.98	+195.0% Llama 4 Maverick is 2.9× more expensive
End-to-end time (lower is better)	49.98s	10.41s	−79.2% Llama 4 Maverick is 79.2% faster
Price per 1M tokens (in / out)	$0.03 / $0.12	$0.25 / $0.92
Input modalities	Text	Text, Image

Llama 4 Maverick vs Granite 4.2 3B: 2.2% more intelligence, 195.0% more cost to run the AA benchmark, and 79.2% faster end-to-end.

Open this matchup in the interactive picker Granite 4.2 3B: what beats it Llama 4 Maverick: what beats it

Today's Pareto frontier

Against the frontier

Granite 4.2 3B is off the frontier: at the same or lower cost, GPT-6 Luna (Low) scores 21.50 for $10.6.
Llama 4 Maverick is off the frontier: at the same or lower cost, GPT-6 Luna (Medium) scores 29.90 for $31.2.

Granite 4.2 3BLlama 4 MaverickToday'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/granite-4-2-3b/llama-4-maverick · Markdown: /api/v1/compare/granite-4-2-3b/llama-4-maverick?format=md · llms.txt · agent skill