DeepSeek-V3

Open DeepSeek MoE chat model for coding, math, and general reasoning

ECI STANDING#143 of 266
SCORE132.35128.55–135.57 CI
FRONTIER LAGLaunched 3.4 months behindScores about what o1-mini did in September 2024.
SCORES LIKE THESE MODELS
Qwen2.5-MaxAlibaba · Jan 2025132.53
Llama 4 MaverickMeta AI · Apr 2025132.2
Mistral Small 3.2Mistral AI · Jun 2025131.74
Gemini 1.5 Pro (Sept 2024)Google DeepMind · Sep 2024131.73
Full Epoch Capabilities Index ranking

From the lab

Launch announcement · Dec 26, 2024🚀 Introducing DeepSeek-V3
RELEASE DATEDec 26, 2024
WEIGHTSOpen weights
CONTEXT WINDOW131.1K
MAX OUTPUT8.2K
REASONINGNo
TOOL CALLINGYes
INPUTStext
OUTPUTStext

Specifications from Models.dev. Availability and limits may differ by provider. Benchmark scores come from Epoch AI under CC BY 4.0.