Gemma 3 1B Instruct vs Llama 3.2 Instruct 1B
Wins 1 of 5 areas
Reasoning
Wins 3 of 5 areas
Facts · Long documents · Following instructions
Llama 3.2 Instruct 1B is the stronger all-rounder.Gemma 3 1B Instruct is better at reasoning.
Scores updated · 32 tests both models report · How we compare
Where each one wins
Tests won in each of the five areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- FactsGetting facts right instead of making them up02Llama 3.2 Instruct 1B2 of 2 tests
- Long documentsFinding answers in very long texts01Llama 3.2 Instruct 1B1 of 1 test
- Following instructionsDoing exactly what it is asked01Llama 3.2 Instruct 1B1 of 1 test
- ReasoningHard problems that need careful thinking10Gemma 3 1B Instruct1 of 3 tests · 2 ties
- CodingWriting and fixing software00Even0 each · 2 ties
Long documents and following instructions rest on a single test each.
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.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Llama 3.2 Instruct 1B pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+16points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+3.2points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+2.9points ahead
Where Gemma 3 1B Instruct pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+4.1points ahead
Every test, side by side
All 32 tests both models report. The winning score is in its model's colour; marks a score checked independently.
ReasoningGemma 3 1B Instruct
- GPQA DiamondGemma 3 1B Instruct by 4.123.719.6+4.1
- Humanity's Last Examtie5.35.5tie
- CritPttie00tie
FactsLlama 3.2 Instruct 1B
- AA-Omniscience · Non-hallucinationLlama 3.2 Instruct 1B by 1617.733.7+16
- AA-Omniscience · AccuracyLlama 3.2 Instruct 1B by 3.23.87+3.2
Long documentsLlama 3.2 Instruct 1B
- AA-LCRLlama 3.2 Instruct 1B by 6.706.7+6.7
Following instructionsLlama 3.2 Instruct 1B
- IFBenchLlama 3.2 Instruct 1B by 2.919.922.8+2.9
Other results23 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- IFEvalGemma 3 1B Instruct by 27.880.252.4+27.8
- GSM8KGemma 3 1B Instruct by 27.162.835.7+27.1
- MATH-500 (EM)Gemma 3 1B Instruct by 21.845.223.4+21.8
- AA-OmniscienceLlama 3.2 Instruct 1B by 20.7-75.5-54.7+20.7
- MGSMGemma 3 1B Instruct by 14.543.629.1+14.5
- HumanEval+Gemma 3 1B Instruct by 1437.223.2+14
- τ²-Bench Telecom (AA run)Gemma 3 1B Instruct by 10.510.50+10.5
- J-GSM8KGemma 3 1B Instruct by 8.433.625.2+8.4
- JGPQALlama 3.2 Instruct 1B by 8.124.232.3+8.1
- JHumanEval+Gemma 3 1B Instruct by 7.32517.7+7.3
- MMLULlama 3.2 Instruct 1B by 6.540.146.6+6.5
- Avg.Gemma 3 1B Instruct by 6.324.618.3+6.3
- MMLU-ProLlama 3.2 Instruct 1B by 6.114.720.8+6.1
- BFCL v3Llama 3.2 Instruct 1B by 4.816.621.4+4.8
- J-MATH500Gemma 3 1B Instruct by 4.215.611.4+4.2
- J-BFCLv3Llama 3.2 Instruct 1B by 3.817.321.1+3.8
- MMMLULlama 3.2 Instruct 1B by 3.834.438.1+3.8
- AA Agentic IndexGemma 3 1B Instruct by 3.53.50+3.5
- Domain AvgGemma 3 1B Instruct by 2.723.921.2+2.7
- JMMLU-ProXLlama 3.2 Instruct 1B by 1.814.115.9+1.8
- LCB v5tie4.43.6tie
- JMMLUtie34.534tie
- AA Intelligencetie4.84.8tie
Questions people ask
Which is better, Gemma 3 1B Instruct or Llama 3.2 Instruct 1B?
Llama 3.2 Instruct 1B wins three of the five areas where both have results: facts, long documents and following instructions. Gemma 3 1B Instruct wins reasoning. They are level on coding.
Which is better for coding?
Neither. All 2 coding tests both models report are ties.
How do you compare the two?
We use the 32 benchmark tests both models have published scores on. The verdict counts the 9 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 23 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.