GPT-5 mini vs Mistral Medium 3.1 (Non-Reasoning)
Wins 7 of 7 areas
Coding · Agents · Reasoning · Facts · Images and charts · Long documents · Following instructions
Wins 0 of 7 areas
—
GPT-5 mini is the stronger all-rounder.
Scores updated · 24 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.
- FactsGetting facts right instead of making them up30GPT-5 mini3 of 3 tests
- AgentsCarrying out multi-step tasks on its own20GPT-5 mini2 of 3 tests · 1 tie
- ReasoningHard problems that need careful thinking20GPT-5 mini2 of 3 tests · 1 tie
- Images and chartsUnderstanding pictures, charts and video20GPT-5 mini2 of 2 tests
- CodingWriting and fixing software21GPT-5 mini2 of 3 tests
- Long documentsFinding answers in very long texts10GPT-5 mini1 of 1 test
- Following instructionsDoing exactly what it is asked10GPT-5 mini1 of 1 test
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.
GPT-5 mini costs 6% 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 GPT-5 mini pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+35.6points ahead
- Reasons across sets of long documentsAA-LCR+29.6points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+28.2points ahead
Where Mistral Medium 3.1 (Non-Reasoning) pulls ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+10.2points 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.
CodingGPT-5 mini
- Terminal-Bench HardGPT-5 mini by 22.733.310.6+22.7
- Terminal-Bench 2.1Mistral Medium 3.1 (Non-Reasoning) by 10.23.713.9+10.2
- SciCodeGPT-5 mini by 6.63932.4+6.6
AgentsGPT-5 mini
- GDPValGPT-5 mini by 13.513.50+13.5
- τ-Bench V3 · BankingGPT-5 mini by 7.315.58.2+7.3
- Terminal-Bench 4.0tie00tie
ReasoningGPT-5 mini
- GPQA DiamondGPT-5 mini by 2482.858.8+24
- Humanity's Last ExamGPT-5 mini by 16.821.54.7+16.8
- CritPttie00tie
FactsGPT-5 mini
- AA-Omniscience · Non-hallucinationGPT-5 mini by 28.243.615.4+28.2
- Vectara HHEM hallucination ratelower is betterGPT-5 mini by 9.812.922.7+9.8
- AA-Omniscience · AccuracyGPT-5 mini by 4.22520.8+4.2
Images and chartsGPT-5 mini
- MMMU-ProGPT-5 mini by 15.970.154.2+15.9
- LMArena · VisionGPT-5 mini by 30 rating points12021172+30 rating
Following instructionsGPT-5 mini
- IFBenchGPT-5 mini by 35.675.439.8+35.6
Other results8 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.
- AA-OmniscienceGPT-5 mini by 29-17.3-46.3+29
- τ²-Bench Telecom (AA run)GPT-5 mini by 27.868.440.6+27.8
- vectara_factual_consistencyGPT-5 mini by 9.887.177.3+9.8
- AA IntelligenceGPT-5 mini by 7.616.89.2+7.6
- AA Agentic IndexGPT-5 mini by 5.88.93.1+5.8
- Artificial Analysis Coding IndexMistral Medium 3.1 (Non-Reasoning) by 4.915.620.5+4.9
- vectara_avg_summary_lengthGPT-5 mini by 26.8 rating points169.7142.9+26.8 rating
- vectara_answer_ratetie99.999.7tie
Questions people ask
Which is better, GPT-5 mini or Mistral Medium 3.1 (Non-Reasoning)?
GPT-5 mini wins all seven areas where both have results: coding, agents, reasoning, facts, images and charts, long documents and following instructions. Mistral Medium 3.1 (Non-Reasoning) wins none.
Which is better for coding?
GPT-5 mini. It wins 2 of the 3 coding tests both models report; Mistral Medium 3.1 (Non-Reasoning) wins 1.
Which is cheaper?
GPT-5 mini costs $0.25 per million input tokens and $2.00 per million output tokens; Mistral Medium 3.1 (Non-Reasoning) costs $0.40 and $2.00. That makes GPT-5 mini about 6% 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 16 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 8 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.