DeepSeek-V3 vs Llama 3.2 3B Instruct
Wins 4 of 5 areas
Coding · Reasoning · Long documents · Following instructions
Wins 0 of 5 areas
—
DeepSeek-V3 is the stronger all-rounder.Llama 3.2 3B Instruct is cheaper.
Scores updated · 24 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.
- CodingWriting and fixing software20DeepSeek-V32 of 2 tests
- ReasoningHard problems that need careful thinking10DeepSeek-V31 of 2 tests · 1 tie
- Long documentsFinding answers in very long texts10DeepSeek-V31 of 1 test
- Following instructionsDoing exactly what it is asked10DeepSeek-V31 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
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.
Llama 3.2 3B Instruct costs 67% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where DeepSeek-V3 pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+40points ahead
- Reasons across sets of long documentsAA-LCR+36.4points ahead
- Recent programming contest problemsLiveCodeBench v6+35.4points ahead
Where Llama 3.2 3B Instruct pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+28.4points ahead
Every test, side by side
All 24 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingDeepSeek-V3
- LiveCodeBench v6DeepSeek-V3 by 35.446.911.5+35.4
- SciCodeDeepSeek-V3 by 33.8395.2+33.8
ReasoningDeepSeek-V3
- GPQA DiamondDeepSeek-V3 by 4065.525.5+40
- Humanity's Last Examtie4.75.3tie
FactsEven
- AA-Omniscience · Non-hallucinationLlama 3.2 3B Instruct by 28.414.142.5+28.4
- AA-Omniscience · AccuracyDeepSeek-V3 by 19.225.46.3+19.2
Following instructionsDeepSeek-V3
- IFBenchDeepSeek-V3 by 14.84126.2+14.8
Other results16 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.
- MATHDeepSeek-V3 by 81.790.28.5+81.7
- MATH-500 (EM)DeepSeek-V3 by 4990.241.2+49
- MMLU-ProDeepSeek-V3 by 40.975.935+40.9
- BBHDeepSeek-V3 by 40.787.546.8+40.7
- HumanEvalDeepSeek-V3 by 37.265.228+37.2
- MMMLUDeepSeek-V3 by 31.579.447.9+31.5
- MMLUDeepSeek-V3 by 28.188.560.4+28.1
- MBPPDeepSeek-V3 by 26.775.448.7+26.7
- τ²-Bench Telecom (AA run)DeepSeek-V3 by 2647.121.1+26
- MGSMDeepSeek-V3 by 21.679.858.2+21.6
- HellaSwagDeepSeek-V3 by 19.188.969.8+19.1
- ARC-ChallengeDeepSeek-V3 by 16.795.378.6+16.7
- GSM8KDeepSeek-V3 by 16.39477.7+16.3
- IFEvalDeepSeek-V3 by 8.786.177.4+8.7
- AA-OmniscienceDeepSeek-V3 by 6.8-40.7-47.5+6.8
- AA IntelligenceDeepSeek-V3 by 49.75.7+4
Questions people ask
Which is better, DeepSeek-V3 or Llama 3.2 3B Instruct?
DeepSeek-V3 wins four of the five areas where both have results: coding, reasoning, long documents and following instructions. Llama 3.2 3B Instruct wins none, but costs 67% less. They are level on facts.
Which is better for coding?
DeepSeek-V3. It wins 2 of the 2 coding tests both models report; Llama 3.2 3B Instruct wins none.
Which is cheaper?
DeepSeek-V3 costs $0.24 per million input tokens and $0.90 per million output tokens; Llama 3.2 3B Instruct costs $0.05 and $0.33. That makes Llama 3.2 3B Instruct about 67% cheaper for the same work.
How do you compare the two?
We use the 24 benchmark tests both models have published scores on. The verdict counts the 8 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 16 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.