Claude Fable 5$22.000/MClaude Opus 4.8$11.000/MClaude Opus 4.7$11.000/MClaude Opus 4.6$11.000/MClaude Opus 4.5$33.000/MClaude Sonnet 3.7$6.600/MClaude Opus 3$33.000/MClaude 2.1$12.800/MClaude 2$12.800/MGPT-5.5$12.500/MGPT-5.2$5.425/MGPT-5.2-Codex$5.425/MGPT-5$3.875/MGPT-4.5$97.500/MGPT-4 Turbo Preview$16.000/MGPT-4$39.000/MGPT-4-32k$78.000/Mo3$19.000/Mo3-mini$2.090/Mo4-mini$2.090/Mo1$28.500/Mo1-mini$5.700/Mo1-preview$28.500/MGemini 3.5 Pro$5.000/MGemini 3.1 Pro$5.000/MGemini 3 Pro$5.000/MGemini 2.5 Pro$3.875/MGemini 1.5 Pro$2.375/MGemini 1.0 Ultra$12.000/MGemini 1.0 Pro$0.800/MClaude Fable 5$22.000/MClaude Opus 4.8$11.000/MClaude Opus 4.7$11.000/MClaude Opus 4.6$11.000/MClaude Opus 4.5$33.000/MClaude Sonnet 3.7$6.600/MClaude Opus 3$33.000/MClaude 2.1$12.800/MClaude 2$12.800/MGPT-5.5$12.500/MGPT-5.2$5.425/MGPT-5.2-Codex$5.425/MGPT-5$3.875/MGPT-4.5$97.500/MGPT-4 Turbo Preview$16.000/MGPT-4$39.000/MGPT-4-32k$78.000/Mo3$19.000/Mo3-mini$2.090/Mo4-mini$2.090/Mo1$28.500/Mo1-mini$5.700/Mo1-preview$28.500/MGemini 3.5 Pro$5.000/MGemini 3.1 Pro$5.000/MGemini 3 Pro$5.000/MGemini 2.5 Pro$3.875/MGemini 1.5 Pro$2.375/MGemini 1.0 Ultra$12.000/MGemini 1.0 Pro$0.800/M
BETA
Subscribe
← Back to Index·All DeepSeek models
DeepSeek
DeepSeekReasoning

DeepSeek-R1-Distill-32B

Distilled

R1 distilled into Qwen 32B. A strong middle ground for self-hosted reasoning that fits on a single high-end GPU.

DeepSeek-R1-Distill-32B is a reasoning AI model from DeepSeek. It costs $0.200 per million input tokens and $0.500 per million output tokens (blended $0.290/M), with a 128K-token context window.

Released Jan 2025Modalities textOfficial model page ↗Provider pricing ↗API docs ↗Compare with another model →Estimate monthly cost →
INPUT
$0.200/M
per million input tokens
OUTPUT
$0.500/M
per million output tokens
BLENDED 70/30
$0.290/M
default reference rate · how it's calculated →
CONTEXT
128K
128,000 tokens
What it's good at
  • Single-GPU reasoning
  • Open weights
  • Strong math
Typical use cases
  • On-prem reasoning
  • Reasoning research
Benchmarks
vs. best public score
MMLU80%
Multitask academic knowledge across 57 subjects.
GPQA Diamond62%
Graduate-level science questions, "Google-proof".
MATH90%
High-school competition math problems.
HumanEval80%
Python function synthesis from docstrings.
Hand-curated from each provider's published reports and public leaderboards. Methodology varies across sources — treat as directional rather than authoritative.
More from DeepSeek
See all 11 →
DeepSeek-V4-Pro
Frontier · 1M ctx
in $1.740/Mout $3.480/M
DeepSeek-V4-Flash
Efficient · 1M ctx
in $0.140/Mout $0.280/M
DeepSeek-V3
Frontier · 64K ctx
in $0.270/Mout $1.100/M
DeepSeek-V3-0324
Frontier · 128K ctx
in $0.270/Mout $1.100/M
DeepSeek-R1
Reasoning · 64K ctx
in $0.550/Mout $2.190/M
DeepSeek-R1-0528
Reasoning · 64K ctx
in $0.550/Mout $2.190/M
Frequently asked questions

How much does DeepSeek-R1-Distill-32B cost?

DeepSeek-R1-Distill-32B costs $0.200 per million input tokens and $0.500 per million output tokens, for a blended reference rate of $0.290 per million tokens.

What is DeepSeek-R1-Distill-32B's context window?

DeepSeek-R1-Distill-32B supports up to 128K tokens of context (128,000 tokens).

What is DeepSeek-R1-Distill-32B best for?

DeepSeek-R1-Distill-32B is well suited to Single-GPU reasoning, Open weights and Strong math.

Who makes DeepSeek-R1-Distill-32B?

DeepSeek-R1-Distill-32B is developed and served by DeepSeek. It was released in Jan 2025.