DeepSeek-R1 vs Magistral Medium 1
Wins 2 of 6 areas
Long documents · Following instructions
Wins 1 of 6 areas
Agents
DeepSeek-R1 wins more areas, narrowly.Magistral Medium 1 is better at agents.
Scores updated · 19 tests both models report · How we compare
Where each one wins
Tests won in each of the six 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.
- Long documentsFinding answers in very long texts10DeepSeek-R11 of 1 test
- Following instructionsDoing exactly what it is asked10DeepSeek-R11 of 1 test
- AgentsCarrying out multi-step tasks on its own01Magistral Medium 11 of 1 test
- CodingWriting and fixing software11Even1 each
- ReasoningHard problems that need careful thinking11Even1 each · 1 tie
- FactsGetting facts right instead of making them up11Even1 each
Agents, 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 three tests each model wins by the widest margin. Scores are out of 100.
Where DeepSeek-R1 pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+13.9points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+10.2points ahead
- Code for real scientific research problemsSciCode+8.6points ahead
Where Magistral Medium 1 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+31points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+3points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+1.3points ahead
Every test, side by side
All 19 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingEven
- SciCodeDeepSeek-R1 by 8.638.329.7+8.6
- Terminal-Bench HardMagistral Medium 1 by 36.19.1+3
ReasoningEven
- GPQA DiamondDeepSeek-R1 by 2.970.867.9+2.9
- Humanity's Last ExamMagistral Medium 1 by 1.38.59.8+1.3
- CritPttie0.60.3tie
FactsEven
- AA-Omniscience · Non-hallucinationMagistral Medium 1 by 319.740.7+31
- AA-Omniscience · AccuracyDeepSeek-R1 by 10.230.520.3+10.2
Following instructionsDeepSeek-R1
- IFBenchDeepSeek-R1 by 13.93925.1+13.9
Other results9 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.
- LiveCodeBenchDeepSeek-R1 by 13.263.550.3+13.2
- τ²-Bench Telecom (AA run)Magistral Medium 1 by 11.711.423.1+11.7
- ARC-AGI-1DeepSeek-R1 by 9.715.86.1+9.7
- AIME 2024DeepSeek-R1 by 9.179.870.7+9.1
- Artificial Analysis Coding IndexDeepSeek-R1 by 8.624.616+8.6
- Aider-PolyglotDeepSeek-R1 by 6.253.347.1+6.2
- AA-OmniscienceMagistral Medium 1 by 5.3-32.2-26.9+5.3
- AIME 2025DeepSeek-R1 by 5.17064.9+5.1
- AA IntelligenceDeepSeek-R1 by 2.311.49.1+2.3
Questions people ask
Which is better, DeepSeek-R1 or Magistral Medium 1?
DeepSeek-R1 wins two of the six areas where both have results: long documents and following instructions. Magistral Medium 1 wins agents. They are level on coding, reasoning and facts.
Which is better for coding?
Neither. They win 1 coding test each of the 2 both models report.
How do you compare the two?
We use the 19 benchmark tests both models have published scores on. The verdict counts the 10 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 9 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.