Model comparison
Kimi K2 vs DeepSeek-R1
- Input
- $0.55 / 1M tokens
- Output
- $2.20 / 1M tokens
- Context
- 128K
- Released
- 2025-07-11
- Input
- $0.55 / 1M tokens (cache miss; $0.14 cache hit)
- Output
- $2.19 / 1M tokens
- Context
- 128K
- Released
- 2025-01-20
Kimi K2 and DeepSeek-R1 split the 6 benchmarks both report evenly; both charge the same $0.6 per 1M input tokens.
Specs and pricing side by side
| Spec | Kimi K2 | DeepSeek-R1 |
|---|---|---|
| Maker | Moonshot AI (Kimi) | DeepSeek |
| Released | 2025-07-11 | 2025-01-20 |
| Context window | 128K | 128Ksame |
| Max output | 128K | 32,768 tokens (max generation length) |
| Input price | $0.55 / 1M tokens | $0.55 / 1M tokens (cache miss; $0.14 cache hit) |
| Output price | $2.20 / 1M tokens | $2.19 / 1M tokens |
| Parameters | 1T total / 32B active (MoE) | 671B total / 37B active (Mixture-of-Experts) |
| License | Modified MIT License | MIT |
| Weights | Open weights | Open weightssame |
| Modalities | Text | Textsame |
| Knowledge cutoff | not published | Not officially disclosed |
Benchmarks both models report
| Benchmark | Kimi K2 | DeepSeek-R1 |
|---|---|---|
| SWE-bench Verified | 65.8% | 49.2% |
| AIME 2024 | 69.6% | 79.8% |
| MATH-500 | 97.4% | 97.3% |
| MMLU | 89.5% | 90.8% |
| MMLU-Pro | 81.1% | 84% |
| GPQA-Diamond | 75.1% | 71.5% |
Scores are as published by each maker or its cited source; harness, reasoning effort and date can differ between rows, so treat small gaps as ties. Sources are listed on each model page.
Which one should you pick?
Pick Kimi K2 if…
- You care about SWE-bench Verified, MATH-500, GPQA-Diamond (it scores higher on 3 shared benchmarks).
- Your use case is: agentic coding assistants and automated software-engineering workflows.
Pick DeepSeek-R1 if…
- You care about AIME 2024, MMLU, MMLU-Pro (it scores higher on 3 shared benchmarks).
- Your use case is: competition-style math and multi-step logical reasoning.
FAQ
Which is better, Kimi K2 or DeepSeek-R1?
Kimi K2 and DeepSeek-R1 split the 6 benchmarks both report evenly; both charge the same $0.6 per 1M input tokens. The better pick depends on whether you value benchmark scores, price per token, context length or openness for your workload.
How do Kimi K2 and DeepSeek-R1 pricing compare?
Kimi K2: $0.55 / 1M tokens input, $2.20 / 1M tokens output. DeepSeek-R1: $0.55 / 1M tokens (cache miss; $0.14 cache hit) input, $2.19 / 1M tokens output (list API prices as published by each maker).
What are the context windows of Kimi K2 and DeepSeek-R1?
Kimi K2 offers 128K with up to 128K output; DeepSeek-R1 offers 128K with up to 32,768 tokens (max generation length) output.
Is Kimi K2 newer than DeepSeek-R1?
Kimi K2 was released 2025-07-11; DeepSeek-R1 was released 2025-01-20.