The Open-Source AI Stack
RSS
All models

Models · Compare

DeepSeek-R1 (May 2025 refresh) vs Claude Sonnet 4

Rows highlighted in warm gray are where the models differ. Numbers carry their as-of date and primary source.

Specs

Field A: DeepSeek-R1 (May 2025 refresh) B: Claude Sonnet 4
Released 2025-05-282025-05-22
Developer DeepSeekAnthropic
Openness OpenProprietary
License MITProprietary
OSI-approved yesno
Data released nono
Training code nono
Architecture moeunknown
Total params
Active params
Experts
Context window
Attention mlaunknown
Position enc. rope-yarnunknown
Pretraining tokens
Post-training sft, grpo, rejection-samplingrlhf, constitutional
Training hardware H800
$/M input $3.00
$/M output $15.00
Output tok/sec 48.4

Benchmarks

Missing scores render as not reported; never inferred. Bold highlights the leader per benchmark.

General reasoning

MMLU-Pro 85.0 2025-05-28 83.7 2026-05-21
GPQA-Diamond 81.0 2025-05-28 70.0 2025-05-22

Code

SWE-Bench Verified 72.7 2025-05-22
LiveCodeBench 73.3 2025-05-28 44.9 2026-05-21

Math

MATH 93.4 2026-05-21
AIME 2024 91.4 2025-05-28 40.7 2026-05-21
AIME 2025 33.1 2025-05-22

Context · A

An RL-only refresh of R1 that gained substantial ground on reasoning benchmarks (notably AIME 2024) without any new pretraining. Tightened the open-vs-closed reasoning gap.

Context · B

Mid-tier model of the May 22 2025 Claude 4 launch. Inherited the hybrid-reasoning approach from Claude 3.7 Sonnet with near-instant and extended-thinking modes, plus parallel tool execution and an extended-thinking-with-tool-use beta. Held the SWE-Bench Verified lead for closed mid-tier coding through summer 2025 at the same $3 / $15 price point as 3.5 and 3.7 Sonnet.

DeepSeek-R1 (May 2025 refresh) detail → · Claude Sonnet 4 detail →