All AI models · LLM benchmarks · Methodology

DeepSeek

DeepSeek R1 0528 Qwen3 8B

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

Model facts

Context window
128K tokens
Maximum output
32K tokens
Cheapest paid input
$0.06 per 1M tokens
Cheapest paid output
$0.09 per 1M tokens
Open weights
yes
Providers
1

Capabilities and modalities

reasoning, temperature control, open weights, text.

Local hardware estimate

At 4K context: Q4_K_M 6.0 GiB, Q8_0 9.8 GiB, F16 17.0 GiB working memory.

Parameters
8.2 billion
Architecture
qwen3
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
default10.3not measured63.7not measured

Observed price history

  • 2026-08-24: $0.06 input / $0.09 output per 1M tokens

API providers

  • NovitaAI — model id deepseek/deepseek-r1-0528-qwen3-8b; $0.06 input / $0.09 output per 1M tokens; provider documentation