DeepSeek-R1-Distill-Qwen-1.5B vs DeepSeek-R1-Distill-Qwen-14B
Wins 0 of 4 areas
—
Wins 4 of 4 areas
Coding · Reasoning · Long documents · Following instructions
DeepSeek-R1-Distill-Qwen-14B is the stronger all-rounder.
Scores updated · 13 tests both models report · How we compare
Where each one wins
Tests won in each of the four 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.
- ReasoningHard problems that need careful thinking02DeepSeek-R1-Distill-Qwen-14B2 of 2 tests
- CodingWriting and fixing software01DeepSeek-R1-Distill-Qwen-14B1 of 1 test
- Long documentsFinding answers in very long texts01DeepSeek-R1-Distill-Qwen-14B1 of 1 test
- Following instructionsDoing exactly what it is asked01DeepSeek-R1-Distill-Qwen-14B1 of 1 test
Coding, 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 DeepSeek-R1-Distill-Qwen-14B pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+38.6points ahead
- Code for real scientific research problemsSciCode+17.3points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+8.9points ahead
Where DeepSeek-R1-Distill-Qwen-1.5B pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 13 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingDeepSeek-R1-Distill-Qwen-14B
- SciCodeDeepSeek-R1-Distill-Qwen-14B by 17.36.623.9+17.3
ReasoningDeepSeek-R1-Distill-Qwen-14B
- GPQA DiamondDeepSeek-R1-Distill-Qwen-14B by 38.69.848.4+38.6
- Humanity's Last ExamDeepSeek-R1-Distill-Qwen-14B by 13.14.1+1
Long documentsDeepSeek-R1-Distill-Qwen-14B
- AA-LCRDeepSeek-R1-Distill-Qwen-14B by 8.61.710.3+8.6
Following instructionsDeepSeek-R1-Distill-Qwen-14B
- IFBenchDeepSeek-R1-Distill-Qwen-14B by 8.913.222.1+8.9
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.
- Codeforces (Rating)DeepSeek-R1-Distill-Qwen-14B by 527 rating points9541481+527 rating
- MMLU-ProDeepSeek-R1-Distill-Qwen-14B by 41.82768.8+41.8
- AIME 2024DeepSeek-R1-Distill-Qwen-14B by 40.828.969.7+40.8
- LiveCodeBenchDeepSeek-R1-Distill-Qwen-14B by 36.216.953.1+36.2
- AIME 2025DeepSeek-R1-Distill-Qwen-14B by 29.22049.2+29.2
- HMMT Feb. 2025DeepSeek-R1-Distill-Qwen-14B by 2011.731.7+20
- MATH-500 (EM)DeepSeek-R1-Distill-Qwen-14B by 1083.993.9+10
- AA IntelligenceDeepSeek-R1-Distill-Qwen-14B by 2.35.57.8+2.3
Questions people ask
Which is better, DeepSeek-R1-Distill-Qwen-1.5B or DeepSeek-R1-Distill-Qwen-14B?
DeepSeek-R1-Distill-Qwen-14B wins all four areas where both have results: coding, reasoning, long documents and following instructions. DeepSeek-R1-Distill-Qwen-1.5B wins none.
Which is better for coding?
DeepSeek-R1-Distill-Qwen-14B. It wins the one coding test both models report.
How do you compare the two?
We use the 13 benchmark tests both models have published scores on. The verdict counts the 5 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.