LLM ReferenceLLM Reference
All comparisons

GPT-4 Turbo vs DeepSeek V3

Side-by-side comparison of specifications, capabilities, and pricing.

Released2024-04-092024-12-26
Context window128K64k
Parameters1.76T (8x222B MoE)*671B
Architecturemixture of expertsmixture of experts
LicenseProprietaryOpen Source
Knowledge cutoff2023-122024-04

Capabilities

Vision
Multimodal
Reasoning
Function calling
Tool use
Structured Outputs
Code execution

Availability

Providers

Benchmarks

Massive Multitask Language Understanding86.588.5Chatbot Arena1242.01302.0MMLU PRO69.475.9HellaSwag—95.7HumanEval—85.5LiveCodeBench—49.6Aider Polyglot—48.4BigCodeBench—50.0