localai.computer

Models GPUs Systems AI Setups Builds Methodology

Loading content...

Resources

Methodology
Submit Benchmark
About

Browse

AI Models
GPUs
PC Builds

Community

Leaderboard

Legal

Privacy
Terms
Contact

© 2025 localai.computer. Hardware recommendations for running AI models locally.

We earn from qualifying purchases through affiliate links at no extra cost to you. This supports our free content and research.

Loading developer profiles...

Loading developer profile...

Home
Home
Developers
deepseek-ai

deepseek-ai

deepseek-ai publishes models from 3B to 685B parameters focused on workstation-grade inference.

16 models3B to 685B parametersFrom $999

Models

Search and sort every tracked release before opening a dedicated requirements page.

deepseek-ai/DeepSeek-Math-V2

685B

Minimum VRAM: Q4 · 383GB

Throughput: 125.8 tok/s

From $329View requirements →

deepseek-ai/DeepSeek-OCR

3B

Minimum VRAM: Q4 · 2GB

Throughput: 931.5 tok/s

From $329View requirements →

deepseek-ai/deepseek-coder-33b-instruct

33B

Minimum VRAM: Q4 · 17GB

Throughput: 261.8 tok/s

From $329View requirements →

deepseek-ai/DeepSeek-Coder-V2-Instruct-0724

236B

Minimum VRAM: Q4 · 115GB

Throughput: 114.5 tok/s

From $329View requirements →

deepseek-ai/DeepSeek-V2.5

671B

Minimum VRAM: Q4 · 328GB

Throughput: 267.5 tok/s

From $329View requirements →

deepseek-ai/DeepSeek-V3

7B

Minimum VRAM: Q4 · 4GB

Throughput: 784.6 tok/s

From $329View requirements →

deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct

7B

Minimum VRAM: Q4 · 4GB

Throughput: 827.8 tok/s

From $329View requirements →

deepseek-ai/DeepSeek-V3-0324

7B

Minimum VRAM: Q4 · 4GB

Throughput: 799.2 tok/s

From $329View requirements →

deepseek-ai/DeepSeek-R1

7B

Minimum VRAM: Q4 · 4GB

Throughput: 804.6 tok/s

From $329View requirements →

deepseek-ai/DeepSeek-R1-0528

7B

Minimum VRAM: Q4 · 4GB

Throughput: 781.0 tok/s

From $329View requirements →

deepseek-ai/DeepSeek-V3.1

7B

Minimum VRAM: Q4 · 4GB

Throughput: 728.2 tok/s

From $329View requirements →

deepseek-ai/deepseek-coder-1.3b-instruct

3B

Minimum VRAM: Q4 · 2GB

Throughput: 921.1 tok/s

From $329View requirements →

deepseek-ai/DeepSeek-R1-Distill-Llama-8B

8B

Minimum VRAM: Q4 · 4GB

Throughput: 797.1 tok/s

From $329View requirements →

deepseek-ai/DeepSeek-R1-Distill-Qwen-7B

7B

Minimum VRAM: Q4 · 4GB

Throughput: 832.3 tok/s

From $329View requirements →

deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B

5B

Minimum VRAM: Q4 · 3GB

Throughput: 811.7 tok/s

From $329View requirements →

deepseek-ai/DeepSeek-R1-Distill-Qwen-32B

32B

Minimum VRAM: Q4 · 16GB

Throughput: 255.7 tok/s

From $329View requirements →

Recommended hardware

Budget$300-$700

AMD Instinct MI300X • Pricing soon

4GB VRAM target • Q4

Edge inference, personal copilots, prototype agents.

Mid-range$700-$2,000

RTX 4090 • $1,599

7GB VRAM target • Q8

Team chatbots, offline knowledge bases, enterprise POCs.

High-end$2,000+

RTX 4080 Super • $999

15GB VRAM target • FP16

Agent orchestration, evaluation pipelines, long-context tasks.

deepseek-ai AI Models & GPU Requirements - localai.computer