DeepSeek-V3-0324

Mixture-of-Experts model challenging top AI models at much lower cost. Updated on March 24th, 2025.

About model

DeepSeek-V3-0324 is a strong Mixture-of-Experts language model with 671B parameters, 37B activated per token, designed for efficient inference and cost-effective training. It excels in performance, outpacing other open-source models and rivaling leading closed-source models. Suitable for applications requiring high-quality language understanding and generation.

This endpoint was updated on March 24th, 2025 to use the weights of the improved DeepSeek-V3-0324 model.

Quickstart guides

RAG

Building a RAG Workflow

Agents

Agent Workflows

Apps

Next.js Chat Quickstart

Performance benchmarks

Model	GPQA Diamond	HLE	LiveCodeBench	MATH500	SWE-bench verified
DeepSeek-V3-0324	35.3%			56.3%
Related open-source models
Competitor closed-source models
Claude Opus 4.6	90.5%	34.2%			78.7%
OpenAI o3	83.3%	24.9%		99.2%	62.3%
OpenAI o1	76.8%			96.4%	48.9%
GPT-4o	49.2%	2.7%	32.3%	89.3%	31.0%

Related models

Model specifications

Model data

Model provider
DeepSeek
Type
LLM
Main use cases
Chat
Function Calling
Features
Function Calling
JSON Mode
Fine tuning
Supported
Deployment
On-Demand Dedicated
Monthly Reserved
Parameters
671B
Context length
131K
Input price
$1.25 / 1M tokens
Output price
$1.25 / 1M tokens
Input modalities
Text
Output modalities
Text

Released
December 25, 2024
Last updated
January 22, 2026
Quantization level
FP8
External link
Provider docs
Category
Chat

Quickstart docs

Deploy model