Skip to content
VECTOR WIREAI INTELLIGENCE
UTC

Granite 3.2 2B Instruct vs Granite 3.3 8B Instruct

IBM · released

IBM · released

Scores updated · 13 tests both models report · How we compare

What it costs

Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.

Granite 3.2 2B InstructNo current price is tracked.
Granite 3.3 8B Instruct$0.03 to read · $0.25 to write$0.28Price from Artificial Analysis

Every test, side by side

All 13 tests both models report. The winning score is in its model's colour; marks a score checked independently.

Other results13 tests
  • DROPGranite 3.3 8B Instruct by 35.623.859.4+35.6
  • MATH-500 (EM)Granite 3.3 8B Instruct by 33.535.569+33.5
  • Arena HardGranite 3.3 8B Instruct by 32.724.957.6+32.7
  • AlpacaEval-2.0Granite 3.3 8B Instruct by 28.234.562.7+28.2
  • GSM8KGranite 3.3 8B Instruct by 13.96780.9+13.9
  • IFEvalGranite 3.3 8B Instruct by 13.261.574.8+13.2
  • HumanEval+Granite 3.3 8B Instruct by 12.773.486.1+12.7
  • HumanEvalGranite 3.3 8B Instruct by 9.680.189.7+9.6
  • MMLUGranite 3.3 8B Instruct by 8.357.265.5+8.3
  • BigBenchHardGranite 3.3 8B Instruct by 7.761.469.1+7.7
  • TruthfulQAGranite 3.3 8B Instruct by 7.159.866.9+7.1
  • PopQAGranite 3.3 8B Instruct by 5.620.626.2+5.6
  • AttaQGranite 3.3 8B Instruct by 5.383.288.5+5.3

Questions people ask

How do you compare the two?

We use the 13 benchmark tests both models have published scores on. None of them is in the eight capability areas we count, so this page lists them without a verdict. Each score is the one shown on the model's own page.