Logo
together/deepseek-ai/DeepSeek-V3 DeepSeek V3

Released Date: 12/26/2024

Avg. Accuracy:

57.8%

Latency:

7.85s

Performance by Benchmark

Benchmarks

Accuracy

Rankings

ContractLaw

57.8%

( 44 / 48 )

Academic Benchmarks
Proprietary Benchmarks (contact us to get access)

Cost Analysis

Input Cost

$1.25 / M Tokens

Output Cost

$1.25 / M Tokens

Input Cost (per char)

$0.26 / M chars

Output Cost (per char)

$0.25 / M chars

Overview

DeepSeek v3 is the latest generation in Deepseek’s language models, and serves as the base model for their “R1” model. It is open-source.

Key Specifications

  • Context Window: 131,000 tokens
  • Number of Parameters: 671B
  • Pricing: $1.25 / 1M tokens (input and output)
Join our mailing list to receive benchmark updates on

Stay up to date as new benchmarks and models are released.