NEWHow AI models draw the world map · Gecko Tests
Beta
Home/Models/DeepSeek-Coder-V2-Lite-Base
DeepSeek logo

DeepSeek-Coder-V2-Lite-Base

by DeepSeek · Released Jan 2024

Open Source
57.4
avg score
Rank #103
Compare
Better than 67% of all models
Context
N/A
Input $/1M
n/a
Output $/1M
n/a
Type
text
License
Open Source
Benchmarks
4 tested
Data as of
About

Tested on 4 benchmarks · BenchGecko score 57.4. Top scores: GSM8K (67.1%), MMLU (47.3%), Winogrande (45.8%).

Capabilities
math
67.1
#55 globally
knowledge
45.4
#162 globally
Benchmark Scores
Compare All
Tested on 4 benchmarks · Ranked across 2 categories
Score Distribution (all 312 models)
0255075100
▲ You are here
mathCompare math →
GSM8K

Grade school math word problems. 8,500 problems testing multi-step arithmetic reasoning. A foundational math benchmark.

67.1·
knowledgeCompare knowledge →
MMLU

Massive Multitask Language Understanding. 57 subjects from STEM, humanities, and social sciences. The most widely-cited knowledge benchmark.

47.3·
Winogrande

Commonsense coreference resolution. Tests understanding of pronoun references in ambiguous sentences.

45.8·
ARC AI2

AI2 Reasoning Challenge. Grade-school science questions requiring multi-step reasoning. Easy and Challenge sets test different difficulty levels.

43.1·
Quick compare:moonshotai logovs Kimi K2.5moonshotai logovs Kimi K2 0711writer logovs Palmyra X5
Excellent (85+) Good (70-85) Average (50-70) Below (<50)
Similar Models
moonshotai logo
Kimi K2.5
moonshotai
57.1$0.45/1M
moonshotai logo
Kimi K2 0711
moonshotai
57.7$0.57/1M
writer logo
Palmyra X5
writer
57.0$0.60/1M
Links
Info
DeepSeekPricing explorerDevelopers · API
Research
Technical Report
Documentation
API DocsPlayground
Community
@DeepSeek
Source Code
GitHubHugging Face
BenchGecko API
deepseek-coder-v2-lite-base
Specifications
  • Typetext
  • ContextN/A
  • ReleasedJan 2024
  • LicenseOpen Source
  • Statusbenchmark-only
Available On
DeepSeek logoDeepSeekn/a
Categories
mathknowledge
Learn More
transformeropen weightstokens
Share & Export
Tweet
Related Models
moonshotai logoKimi K2.5
57.1
moonshotai logoKimi K2 0711
57.7
writer logoPalmyra X5
57.0
Alibaba logoQwen2.5 32B Instruct
57.9
DeepSeek-Coder-V2-Lite-Base is an open-source text AI model by DeepSeek, released in January 2024. It has an average benchmark score of 57.4.

Related Models

Kimi K2.5 · moonshotaiKimi K2 0711 · moonshotaiPalmyra X5 · writerQwen2.5 32B Instruct · AlibabaQwen3.8 Max (0902) · Alibaba Qwen

Benchmarks

GSM8KMMLUWinograndeARC AI2

Related Pages

DeepSeek · ProviderDeepSeek · EconomyAll ModelsCompare ModelsPricingDevelopers · API

Key facts · as of 2026-06-22

  • DeepSeek-Coder-V2-Lite-Base by DeepSeek. BenchGecko score 57.4, rank 103 of 312 scored models (normalized average of public benchmark scores).
  • List price n/a input · n/a output per 1M tokens (as of 2026-06-22).

How to cite · data as of 2026-06-22

DeepSeek-Coder-V2-Lite-Base · benchmarks, pricing and providers. BenchGecko, data as of 2026-06-22. https://benchgecko.ai/model/deepseek-coder-v2-lite-base

Credit "Source: BenchGecko" with a link. Prices per provider and Gecko Tests are BenchGecko data (CC BY 4.0); benchmark scores keep their original source, listed in the JSON. JSON · llms.txt · MCP