Key Specifications

Vendordeepseek
Versioncoder-7b
Release Date2024-01-25
Context Window16384 tokens
Input Modalitiestext
Output Modalitiestext
LicenseDeepSeek License
Documentationhttps://api-docs.deepseek.com/

Benchmark Performance

Benchmark Score Unit Evaluated At Notes Source
MMLU 63.6 % 2024-01-25 5-shot view
HUMANEVAL 75.9 pass@1 2024-01-25 view
GSM8K 54.7 % 2024-01-25 0-shot CoT view
MATH 32.2 % 2024-01-25 0-shot CoT view
BBH 59.7 % 2024-01-25 3-shot CoT view
GPQA 23.9 % 2024-01-25 0-shot view
IFEVAL 55.4 % 2024-01-25 prompt_strict view
ARC 86.3 % 2024-01-25 challenge view
MUSR 50.5 % 2024-01-25 0-shot view
WINOGRANDE 79.8 % 2024-01-25 0-shot view

Pricing

Tier Price Currency
Input$0.14 / MtokUSD
Output$0.14 / MtokUSD
Cache Read$0 / MtokUSD
Cache Write$0 / MtokUSD

Source: https://api-docs.deepseek.com/quick_start/pricing · as of 2024-01-25

Compliance

  • Data Residency: CN
  • SOC2: ✗
  • HIPAA: ✗
  • GDPR: ✗
  • ISO 27001: ✗

DeepSeek Coder 7B

Model Overview

DeepSeek Coder 7B 代码专用开源模型, 16K 上下文, 7B 参数, 在 7B 规模上代码生成能力领先。

Core Specifications

Vendor Version Release Date Context Window Input Modalities Output Modalities License
Deepseek coder-7b 2024-01-25 16K text text DeepSeek License

Benchmark Performance

Benchmark Score Unit Notes
MMLU (Massive Multitask Language Understanding) 63.6 % 5-shot
HumanEval 75.9 pass@1
GSM8K (Grade School Math 8K) 54.7 % 0-shot CoT
MATH 32.2 % 0-shot CoT
BBH (BIG-Bench Hard) 59.7 % 3-shot CoT
GPQA 23.9 % 0-shot
IFEval 55.4 % prompt_strict
ARC 86.3 % challenge
MUSR 50.5 % 0-shot
WinoGrande 79.8 % 0-shot

Pricing

Input Output Cache Read Cache Write

per million tokens

Strengths

  • Reliable general-purpose model.

Weaknesses

  • Proprietary, not self-hostable.
  • Context window 16K is limited.

Use Cases

  • Code generation and debugging

References