LLMversus
DeepSeek

DeepSeek V3

Complete specs, pricing, and benchmark data for DeepSeek V3 by DeepSeek. Last verified 2026-04-20.

JSON ModeFunctionsStreaming
Pricing

Input / 1M tokens

$0.259

Output / 1M tokens

$0.420

Cached Input / 1M

$0.135

Context & Output

Context Window

128K

Max Output

8,192

TTFT

220ms

Speed

85 tok/s

Benchmarks

Arena ELO

1280

Coding ELO

1300

Reasoning ELO

1260

HumanEval

92

MMLU

87.5

MATH

78.5

GPQA

59

Try DeepSeek V3 API

Start building with DeepSeek V3

Get API Access→

Price History (Input $/M tokens)

Strengths
  • +Excellent coding performance for its price
  • +Open-source and self-hostable
  • +Very competitive benchmarks
  • +Extremely affordable
Limitations
  • -No vision/multimodal support
  • -API availability can be inconsistent
  • -Smaller context than some competitors

Best For

CodingGeneral PurposeCost SensitiveOpen Source

Compare DeepSeek V3 with...

vs Claude Opus 4vs Gemini 2.5 Provs o3vs DeepSeek R1vs o1vs Qwen 3 235B MoEvs Gemini Experimental 1206vs GPT-4.5vs DeepSeek R1 (Groq)vs Llama 4 Maverickvs DeepSeek R1 (Together)vs Grok 3
Official Pricing Page →
Your ad here