LLM ReferenceLLM Reference
All comparisons

DeepSeek V3 vs Qwen2.5 72B

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

Released2024-12-262024-06-07
Context window64k128K
Parameters671B72.7B
Architecturemixture of expertsdecoder only
LicenseOpen SourceApache 2.0
Knowledge cutoff2024-04

Capabilities

Vision
Multimodal
Reasoning
Function calling
Tool use
Structured Outputs
Code execution

Availability

Providers

Benchmarks

HellaSwag95.795.6HumanEval85.592.7Massive Multitask Language Understanding88.588.2LiveCodeBench49.6—Aider Polyglot48.4—BigCodeBench50.0—Chatbot Arena1302.0—MMLU PRO75.9—Google-Proof Q&A—65.4