Back to models

DeepSeek-R1-Distill-Qwen-14B

ECI#139
135
DeepSeek
Jan. 20, 2025
Open weights

DeepSeek-R1-Distill-Qwen-14B is an open-weights model from DeepSeek, released January 20, 2025. It has an ECI score of 135 and ranks 139th of 257 tracked models. At the time of its release, it ranked 5th of 105 tracked models.

DeepSeek-R1-Distill-Qwen-14B vs select alternatives

DomainBenchmark
DeepSeek-R1-Distill-Qwen-14B
DeepSeek
Jan. 20, 2025
Open weights
DeepSeek-V3
DeepSeek
Dec. 26, 2024
Open weights
o1
OpenAI
Dec. 17, 2024
Closed weights
Gemini 2.0 Flash (Dec 2024)
Google DeepMind
Dec. 11, 2024
Closed weights
ECI scoreBest score 167
135132142135
Ranking#139#151#106#144
Games
Chess PuzzlesBest score 72%
1%
n/a
15%
n/a
Mathematics
MATH Level 5Best score 98%
87%
65%
95%
n/a
OTIS Mock AIME 2024-2025Best score 100%
51%
16%
73%
n/a
Science
GPQA DiamondBest score 96%
45%
57%
77%
n/a

Model metadata

General
OrganizationDeepSeek
Release dateJan. 20, 2025
Model accessibilityOpen weights
CountryChina
Knowledge cutoffUnknown
Provider notesDeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
Pricing
Input price$0.15 / 1M tokens
Output price$0.15 / 1M tokens
Training
Parameters1.5e10
Training computeUnknown
Training hardwareUnknown
Training timeUnknown
Training dataset size2.4e10 datapoints