All AI models · LLM benchmarks · Methodology

DeepSeek

DeepSeek-R1

DeepSeek reasoning model for multi-step analysis, math, coding, and tools

Model facts

Context window
128K tokens
Maximum output
33K tokens
Cheapest paid input
$0.4 per 1M tokens
Cheapest paid output
$1.7 per 1M tokens
Open weights
yes
Providers
19

Capabilities and modalities

reasoning, tool calling, structured output, temperature control, open weights, text.

Local hardware estimate

At 4K context: Q4_K_M 419.9 GiB, Q8_0 738.7 GiB, F16 1336.4 GiB working memory.

Parameters
684.5 billion
Architecture
deepseekv3
Model license
mit

Planning estimate adapted from the MIT-licensed whichllm estimator using metadata from Hugging Face; not a vendor minimum requirement.

Artificial Analysis benchmarks

Compare this model on the full LLM benchmark leaderboard.

ConfigurationAAICodingMathOutput tokens/s
Jan '2518.624.668.0not measured

Observed price history

  • 2026-08-24: $0.4 input / $1.7 output per 1M tokens

API providers