Claude 3.5 Haiku vs Magistral Medium 1.2
Wins 0 of 7 areas
—
Wins 4 of 7 areas
Coding · Reasoning · Images and charts · Long documents
Magistral Medium 1.2 wins more areas, narrowly.
Scores updated · 18 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 software03Magistral Medium 1.23 of 3 tests
- ReasoningHard problems that need careful thinking02Magistral Medium 1.22 of 3 tests · 1 tie
- Images and chartsUnderstanding pictures, charts and video01Magistral Medium 1.21 of 1 test
- Long documentsFinding answers in very long texts01Magistral Medium 1.21 of 1 test
- AgentsCarrying out multi-step tasks on its own00Even0 each · 2 ties
- FactsGetting facts right instead of making them up11Even1 each
- Following instructionsDoing exactly what it is asked00Even0 each · 1 tie
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.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Magistral Medium 1.2 pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+33.1points ahead
- Reasons across sets of long documentsAA-LCR+25.7points ahead
- Harder college exam questions with imagesMMMU-Pro+14.1points ahead
Where Claude 3.5 Haiku pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+18.9points ahead
Every test, side by side
All 18 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingMagistral Medium 1.2
- SciCodeMagistral Medium 1.2 by 11.827.439.2+11.8
- Terminal-Bench HardMagistral Medium 1.2 by 10.62.312.9+10.6
- Terminal-Bench 2.1Magistral Medium 1.2 by 2.310.112.4+2.3
ReasoningMagistral Medium 1.2
- GPQA DiamondMagistral Medium 1.2 by 33.140.873.9+33.1
- Humanity's Last ExamMagistral Medium 1.2 by 6.73.610.3+6.7
- CritPttie00.3tie
FactsEven
- AA-Omniscience · Non-hallucinationClaude 3.5 Haiku by 18.958.940+18.9
- AA-Omniscience · AccuracyMagistral Medium 1.2 by 7.613.220.8+7.6
Images and chartsMagistral Medium 1.2
- MMMU-ProMagistral Medium 1.2 by 14.145.659.7+14.1
Long documentsMagistral Medium 1.2
- AA-LCRMagistral Medium 1.2 by 25.727.353+25.7
Other results5 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.
- τ²-Bench Telecom (AA run)Magistral Medium 1.2 by 27.424.652+27.4
- Artificial Analysis Coding IndexMagistral Medium 1.2 by 5.415.921.3+5.4
- AA-OmniscienceClaude 3.5 Haiku by 4.2-22.5-26.7+4.2
- AA IntelligenceMagistral Medium 1.2 by 2.98.911.8+2.9
- AA Agentic Indextie1.91.9tie
Questions people ask
Which is better, Claude 3.5 Haiku or Magistral Medium 1.2?
Magistral Medium 1.2 wins four of the seven areas where both have results: coding, reasoning, images and charts and long documents. Claude 3.5 Haiku wins none. They are level on agents, facts and following instructions.
Which is better for coding?
Magistral Medium 1.2. It wins 3 of the 3 coding tests both models report; Claude 3.5 Haiku wins none.
How do you compare the two?
We use the 18 benchmark tests both models have published scores on. The verdict counts the 13 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 5 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.