Ministral 3 3B vs Ministral 3 8B
Wins 1 of 7 areas
Agents
Wins 4 of 7 areas
Coding · Images and charts · Long documents · Following instructions
Ministral 3 8B wins more areas, narrowly.Ministral 3 3B is cheaper and better at agents.
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 software03Ministral 3 8B3 of 3 tests
- Images and chartsUnderstanding pictures, charts and video01Ministral 3 8B1 of 1 test
- Long documentsFinding answers in very long texts01Ministral 3 8B1 of 1 test
- Following instructionsDoing exactly what it is asked01Ministral 3 8B1 of 1 test
- AgentsCarrying out multi-step tasks on its own10Ministral 3 3B1 of 3 tests · 2 ties
- ReasoningHard problems that need careful thinking11Even1 each · 1 tie
- 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 33% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Ministral 3 8B pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+11.3points ahead
- Reasons across sets of long documentsAA-LCR+8.7points ahead
- Harder college exam questions with imagesMMMU-Pro+7.9points ahead
Where Ministral 3 3B pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+14points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+1.1points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+1points 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 8B
- SciCodeMinistral 3 8B by 5.415.320.7+5.4
- Terminal-Bench HardMinistral 3 8B by 4.504.5+4.5
- Terminal-Bench 2.1Ministral 3 8B by 4.104.1+4.1
AgentsMinistral 3 3B
- τ-Bench V3 · BankingMinistral 3 3B by 14.73.7+1
- GDPValtie00tie
- Terminal-Bench 4.0tie00tie
ReasoningEven
- GPQA DiamondMinistral 3 8B by 11.335.847.1+11.3
- Humanity's Last ExamMinistral 3 3B by 1.15.44.3+1.1
- CritPttie00tie
FactsEven
- AA-Omniscience · Non-hallucinationMinistral 3 3B by 1419.85.8+14
- AA-Omniscience · AccuracyMinistral 3 8B by 4913+4
Images and chartsMinistral 3 8B
- MMMU-ProMinistral 3 8B by 7.938.146+7.9
Following instructionsMinistral 3 8B
- IFBenchMinistral 3 8B by 2.326.829.1+2.3
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 8B by 20.430.550.9+20.4
- WildBenchMinistral 3 8B by 1056.866.8+10
- TriviaQA 5-shotMinistral 3 8B by 8.959.268.1+8.9
- AIME 2024Ministral 3 8B by 8.577.586+8.5
- AGIEval 5-shotMinistral 3 8B by 851.159.1+8
- LiveCodeBenchMinistral 3 8B by 6.854.861.6+6.8
- AIME 2025Ministral 3 8B by 6.672.178.7+6.6
- MMLU Redux 5-shotMinistral 3 8B by 5.873.579.3+5.8
- MMLUMinistral 3 8B by 5.470.776.1+5.4
- Multilingual MMLUMinistral 3 8B by 5.465.270.6+5.4
- AA-OmniscienceMinistral 3 3B by 4.9-64-69+4.9
- Artificial Analysis Coding IndexMinistral 3 8B by 4.94.89.7+4.9
- τ²-Bench Telecom (AA run)Ministral 3 8B by 1.724.926.6+1.7
- AA Intelligencetie4.85.5tie
- MM MTBenchtie7.88.1tie
Questions people ask
Which is better, Ministral 3 3B or Ministral 3 8B?
Ministral 3 8B wins four of the seven areas where both have results: coding, images and charts, long documents and following instructions. Ministral 3 3B wins agents, and costs 33% less. They are level on reasoning and facts.
Which is better for coding?
Ministral 3 8B. It wins 3 of the 3 coding tests both models report; Ministral 3 3B wins none.
Which is cheaper?
Ministral 3 3B costs $0.10 per million input tokens and $0.10 per million output tokens; Ministral 3 8B costs $0.15 and $0.15. That makes Ministral 3 3B about 33% 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.