Ministral 3 14B vs Ministral 3 3B
Wins 6 of 7 areas
Coding · Agents · Reasoning · Images and charts · Long documents · Following instructions
Wins 0 of 7 areas
—
Ministral 3 14B is the stronger all-rounder.Ministral 3 3B is cheaper.
Scores updated · 29 tests both models report · How we compare
Where each one wins
Tests won in each of the seven 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 software30Ministral 3 14B3 of 3 tests
- AgentsCarrying out multi-step tasks on its own10Ministral 3 14B1 of 3 tests · 2 ties
- ReasoningHard problems that need careful thinking10Ministral 3 14B1 of 3 tests · 2 ties
- Images and chartsUnderstanding pictures, charts and video10Ministral 3 14B1 of 1 test
- Long documentsFinding answers in very long texts10Ministral 3 14B1 of 1 test
- Following instructionsDoing exactly what it is asked10Ministral 3 14B1 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
Images and charts, 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.
Ministral 3 3B costs 50% 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 Ministral 3 14B pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+21.4points ahead
- Harder college exam questions with imagesMMMU-Pro+11.7points ahead
- Reasons across sets of long documentsAA-LCR+9.3points ahead
Where Ministral 3 3B pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+12.3points ahead
Every test, side by side
All 29 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingMinistral 3 14B
- Terminal-Bench 2.1Ministral 3 14B by 9.79.70+9.7
- SciCodeMinistral 3 14B by 8.523.815.3+8.5
- Terminal-Bench HardMinistral 3 14B by 4.54.50+4.5
AgentsMinistral 3 14B
- τ-Bench V3 · BankingMinistral 3 14B by 1.96.64.7+1.9
- GDPValtie00tie
- Terminal-Bench 4.0tie00tie
ReasoningMinistral 3 14B
- GPQA DiamondMinistral 3 14B by 21.457.235.8+21.4
- Humanity's Last Examtie4.65.4tie
- CritPttie00tie
FactsEven
- AA-Omniscience · Non-hallucinationMinistral 3 3B by 12.37.519.8+12.3
- AA-Omniscience · AccuracyMinistral 3 14B by 4.613.69+4.6
Images and chartsMinistral 3 14B
- MMMU-ProMinistral 3 14B by 11.749.838.1+11.7
Following instructionsMinistral 3 14B
- IFBenchMinistral 3 14B by 5.23226.8+5.2
Other results15 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.
- Arena HardMinistral 3 14B by 24.655.130.5+24.6
- TriviaQA 5-shotMinistral 3 14B by 15.774.959.2+15.7
- AGIEval 5-shotMinistral 3 14B by 13.764.851.1+13.7
- AIME 2025Ministral 3 14B by 12.98572.1+12.9
- AIME 2024Ministral 3 14B by 12.389.877.5+12.3
- WildBenchMinistral 3 14B by 11.768.556.8+11.7
- LiveCodeBenchMinistral 3 14B by 9.864.654.8+9.8
- Artificial Analysis Coding IndexMinistral 3 14B by 9.614.44.8+9.6
- Multilingual MMLUMinistral 3 14B by 974.265.2+9
- MMLUMinistral 3 14B by 8.779.470.7+8.7
- MMLU Redux 5-shotMinistral 3 14B by 8.58273.5+8.5
- AA-OmniscienceMinistral 3 3B by 2.4-66.4-64+2.4
- τ²-Bench Telecom (AA run)Ministral 3 14B by 2.327.224.9+2.3
- AA IntelligenceMinistral 3 14B by 1.264.8+1.2
- MM MTBenchtie8.57.8tie
Questions people ask
Which is better, Ministral 3 14B or Ministral 3 3B?
Ministral 3 14B wins six of the seven areas where both have results: coding, agents, reasoning, images and charts, long documents and following instructions. Ministral 3 3B wins none, but costs 50% less. They are level on facts.
Which is better for coding?
Ministral 3 14B. It wins 3 of the 3 coding tests both models report; Ministral 3 3B wins none.
Which is cheaper?
Ministral 3 14B costs $0.20 per million input tokens and $0.20 per million output tokens; Ministral 3 3B costs $0.10 and $0.10. That makes Ministral 3 3B about 50% cheaper for the same work.
How do you compare the two?
We use the 29 benchmark tests both models have published scores on. The verdict counts the 14 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 15 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.