DeepSeek Models

DeepSeek-V4-Pro

DeepSeek · DeepSeek

Best for: Teams comparing frontier-style open-weight reasoning and coding models against hosted closed models.

DeepSeek R1 Distill Qwen 14B

DeepSeek · DeepSeek

Best for: Reasoning experiments, local-friendly evaluation, and distilled model comparisons.

Local: Use the exact checkpoint and quantization that matches your hardware and latency target.

DeepSeek-R1 Distill 14B

DeepSeek · DeepSeek-R1

Best for: DeepSeek-R1 Distill 14B is a Qwen2.5-based distillation of DeepSeek's R1 reasoning model, sized to bring stronger step-by-step reasoning to hardware that can't run the full-size R1.

Local: Qwen2.5 distillation of R1. Strong reasoning and coding at 14B. Q4_K_M weights measured at 8.99 GB, confirming the 9.0 GB estimate this entry already carried. Q4 tight in 12 GB; comfortable in 16 GB VRAM. On an RTX 3080 (10 GB) 10% of resident bytes spilled to host and generation ran 30.51 tok/s, 36% of the card's bandwidth ceiling (measured 2026-08-07, Ollama 0.32.5).

DeepSeek R1

DeepSeek · DeepSeek

Best for: Reasoning experiments, coding workflows, and evaluating open reasoning models against closed alternatives.

Local: Can be tested locally through smaller distilled or quantized variants; the full model is better suited to server-class hardware.

DeepSeek Coder V2

DeepSeek · DeepSeek

Best for: Developers comparing open coding models for IDE assistants and coding agents.

Local: Can be tested locally with smaller or quantized builds; larger variants need serious GPU memory.

DeepSeek V3

DeepSeek · DeepSeek

Best for: General DeepSeek-family comparisons and assistant workflow evaluation.

Local: Use the exact checkpoint and quantization that matches your hardware and latency target.

DeepSeek Coder V2 Lite

DeepSeek · DeepSeek

Best for: Coding assistant tests where full-size coder models are too heavy.

Local: Use the exact checkpoint and quantization that matches your hardware and latency target.

DeepSeek Coder 33B

DeepSeek · DeepSeek

Best for: Coding-model comparisons and legacy baseline testing.

Local: Use the exact checkpoint and quantization that matches your hardware and latency target.

DeepSeek Coder 6.7B

DeepSeek · DeepSeek

Best for: Local coding experiments and lightweight code generation baselines.

Local: Use the exact checkpoint and quantization that matches your hardware and latency target.

DeepSeek-R1 8B (0528 Qwen3)

DeepSeek · DeepSeek-R1

Best for: DeepSeek-R1 8B (0528 Qwen3) is a distilled reasoning model that compresses DeepSeek's R1-0528 update into an 8B Qwen3-based checkpoint small enough to run on a single consumer GPU.

Local: Qwen3-8B distillation of the R1-0528 update — Ollama's deepseek-r1:8b tag now points at this checkpoint, replacing the original Llama-3 distill. Improved math, coding, and logic over the first-release distills. Thinking chains visible in output. Q4 fits in 8 GB VRAM.

DeepSeek R1 Distill Llama 70B

DeepSeek · DeepSeek

Best for: Reasoning experiments, local-friendly evaluation, and distilled model comparisons.

Local: Use the exact checkpoint and quantization that matches your hardware and latency target.

DeepSeek R1 Distill Llama 8B

DeepSeek · DeepSeek

Best for: Reasoning experiments, local-friendly evaluation, and distilled model comparisons.

Local: Use the exact checkpoint and quantization that matches your hardware and latency target.

DeepSeek R1 Distill Qwen 32B

DeepSeek · DeepSeek

Best for: Reasoning experiments, local-friendly evaluation, and distilled model comparisons.

Local: Use the exact checkpoint and quantization that matches your hardware and latency target.

DeepSeek R1 Distill Qwen 7B

DeepSeek · DeepSeek

Best for: Reasoning experiments, local-friendly evaluation, and distilled model comparisons.

Local: Use the exact checkpoint and quantization that matches your hardware and latency target.

DeepSeek R1 Distill Qwen 1.5B

DeepSeek · DeepSeek

Best for: Reasoning experiments, local-friendly evaluation, and distilled model comparisons.

Local: Use the exact checkpoint and quantization that matches your hardware and latency target.