Compare the listed model data for DeepSeek-R1 (May 28) by DeepSeek and Kimi-K2 Instruct by Moonshot AI. Benchmark and pricing records are not a controlled test and may change.
Data last updated: 2026-10-16. Verify current model details with each provider.
What the listed data says
DeepSeek-R1 (May 28): 128K tokens; Kimi-K2 Instruct: 128K tokens. The listed context-window values are equal. DeepSeek-R1 (May 28) is listed for coding, reasoning, long-context tasks; Kimi-K2 Instruct is listed for coding, reasoning, agentic workflows, long-context tasks. Shared catalog benchmark entries: AIME 2025: DeepSeek-R1 (May 28) 87.5%; Kimi-K2 Instruct 78.4%; GPQA Diamond: DeepSeek-R1 (May 28) 75.7%; Kimi-K2 Instruct 75.3%; SWE-bench Verified: DeepSeek-R1 (May 28) 57.6%; Kimi-K2 Instruct 65.8%; LiveCodeBench: DeepSeek-R1 (May 28) 69.5%; Kimi-K2 Instruct 75.2%. These records are reference data, not a controlled head-to-head test; verify current pricing and specifications with each provider.
Catalog source links
Use these model and provider links to verify current specifications, pricing, and availability.
The first listed parameter quantities are 671B total (37B active via MoE) and 1T total (32B active via MoE); the larger quantity is 1.49× the smaller. Total and active MoE parameters are not directly equivalent.
The catalog release dates are May 28, 2025 and July 2025, about 2 months apart at the precision supplied.
Cost example: 1M input and 1M output tokens
A cost example for 1M input plus 1M output tokens cannot be calculated reliably because separate numeric input and output rates are not listed for both models.
Choose DeepSeek-R1 (May 28) if…
Choose DeepSeek-R1 (May 28) when your task aligns with catalog-listed uses or strengths such as Software engineering tasks requiring real-world code changes, Competition mathematics at AIME/Olympiad level, Drop-in improvement wherever R1 was previously used, Significant improvement over original R1 — SWE-bench 49.2% → 57.6%.
Choose Kimi-K2 Instruct if…
Choose Kimi-K2 Instruct when your task aligns with catalog-listed uses or strengths such as Elite software engineering AI: autonomous code review, bug fixing in production repos, Enterprise agentic systems with multi-tool orchestration, Frontier mathematical reasoning, 65.8% SWE-bench Verified — highest open-model software engineering score ever recorded.
Feature Comparison — DeepSeek-R1 (May 28) vs Kimi-K2 Instruct
Catalog specifications for DeepSeek-R1 (May 28) and Kimi-K2 Instruct
Specification
DeepSeek-R1 (May 28)
Kimi-K2 Instruct
Provider
DeepSeek
Moonshot AI
Parameters
671B total (37B active via MoE)
1T total (32B active via MoE)
Context window
128K tokens
128K tokens
Maximum output
Not listed
Not listed
Catalog pricing
MIT open source. DeepSeek API: ~$0.55/MTok input, $2.19/MTok output
Moonshot API: beta. Research weights released. Third-party: emerging
Listed input/output cost per 1M tokens
input $0.55, output $2.19 per 1M tokens
Separate numeric input/output rates not listed
License
MIT
Modified Apache 2.0 (Kimi K2 Model License)
Coding support
Listed
Listed
Reasoning support
Listed
Listed
Vision input
Not listed
Not listed
Multimodal
Not listed
Not listed
Open weights
Yes
Yes
Benchmark: AIME 2025
87.5%
78.4%
Benchmark: GPQA Diamond
75.7%
75.3%
Benchmark: SWE-bench Verified
57.6%
65.8%
Benchmark: LiveCodeBench
69.5%
75.2%
Figures are transcribed from LITHOVEX model records. Benchmark versions and evaluation setups may differ; pricing and specifications can change. Check each provider's current documentation before making a decision.
Frequently Asked Questions
What is different between DeepSeek-R1 (May 28) and Kimi-K2 Instruct?
DeepSeek-R1 (May 28): 128K tokens; Kimi-K2 Instruct: 128K tokens. The listed context-window values are equal. DeepSeek-R1 (May 28) is listed for coding, reasoning, long-context tasks; Kimi-K2 Instruct is listed for coding, reasoning, agentic workflows, long-context tasks. Shared catalog benchmark entries: AIME 2025: DeepSeek-R1 (May 28) 87.5%; Kimi-K2 Instruct 78.4%; GPQA Diamond: DeepSeek-R1 (May 28) 75.7%; Kimi-K2 Instruct 75.3%; SWE-bench Verified: DeepSeek-R1 (May 28) 57.6%; Kimi-K2 Instruct 65.8%; LiveCodeBench: DeepSeek-R1 (May 28) 69.5%; Kimi-K2 Instruct 75.2%. These records are reference data, not a controlled head-to-head test; verify current pricing and specifications with each provider.
What benchmark values are listed for DeepSeek-R1 (May 28) and Kimi-K2 Instruct?
AIME 2025: DeepSeek-R1 (May 28) 87.5%; Kimi-K2 Instruct 78.4%; GPQA Diamond: DeepSeek-R1 (May 28) 75.7%; Kimi-K2 Instruct 75.3%; SWE-bench Verified: DeepSeek-R1 (May 28) 57.6%; Kimi-K2 Instruct 65.8%; LiveCodeBench: DeepSeek-R1 (May 28) 69.5%; Kimi-K2 Instruct 75.2%. The catalog values may use different versions and evaluation setups, so they are reference points rather than a controlled head-to-head result.
Which model has the larger listed context window?
DeepSeek-R1 (May 28): 128K tokens; Kimi-K2 Instruct: 128K tokens. The listed context-window values are equal.
What are the listed costs per million tokens for DeepSeek-R1 (May 28) and Kimi-K2 Instruct?
DeepSeek-R1 (May 28): input $0.55, output $2.19 per 1M tokens (MIT open source. DeepSeek API: ~$0.55/MTok input, $2.19/MTok output). Kimi-K2 Instruct: Separate numeric input/output rates not listed (Moonshot API: beta. Research weights released. Third-party: emerging). Where separate numeric input and output rates are available, the worked example uses 1M input plus 1M output tokens. Confirm current terms with each provider.
How far apart are the listed release dates for DeepSeek-R1 (May 28) and Kimi-K2 Instruct?
The catalog lists DeepSeek-R1 (May 28) as May 28, 2025 and Kimi-K2 Instruct as July 2025, about 2 months apart at the precision supplied.
When was the comparison data last updated?
The model catalog records 2026-10-16 as its data update date. Provider specifications and prices may have changed since then.
How should I choose between DeepSeek-R1 (May 28) and Kimi-K2 Instruct?
Compare the listed specifications and capabilities, review the benchmark notes, then test both with the same representative prompt for your task. This page does not claim a universal winner.
Create your free LITHOVEX account and compare DeepSeek-R1 (May 28) and Kimi-K2 Instruct on your actual tasks. No credit card, no API key, no subscriptions needed.