DeepSeek-V3.2 Base vs Kimi K1.5 Chat: which model should you choose?
We compared DeepSeek-V3.2 Base and Kimi K1.5 Chat on price, context window and 12 benchmarks. DeepSeek-V3.2 Base and Kimi K1.5 Chat are effectively tied on the index average; Kimi K1.5 Chat is cheaper on a blended 3:1 basis.
DeepSeek-V3.2 Base
DeepSeek-V3.2 Base is a multimodal model from DeepSeek. It offers a 128,000-token context window, costs $5.05 per million input tokens and $27.10 per million output tokens, and averages 80.2 across the index's benchmark suite.
- Input
- $5.05 / 1M
- Output
- $27.10 / 1M
- Context
- 128,000
- Avg score
- 80.2
- Open weights
- Yes
Kimi K1.5 Chat
Kimi K1.5 Chat is a multimodal model from Moonshot AI. It offers a 128,000-token context window, costs $5.11 per million input tokens and $26.59 per million output tokens, and averages 80.2 across the index's benchmark suite.
- Input
- $5.11 / 1M
- Output
- $26.59 / 1M
- Context
- 128,000
- Avg score
- 80.2
- Open weights
- Yes
Specification comparison
| Spec | DeepSeek-V3.2 Base | Kimi K1.5 Chat | Edge |
|---|---|---|---|
| Input price / 1M | $5.05 | $5.11 | A |
| Output price / 1M | $27.10 | $26.59 | B |
| Context window | 128,000 | 128,000 | Tie |
| Max output | 128,000 | 32,768 | A |
| Average score | 80.2 | 80.2 | Tie |
| Arena Elo | 1417.9 | 1422.6 | B |
| Open weights | Yes | Yes | Tie |
Benchmark comparison
DeepSeek-V3.2 Base wins 5, Kimi K1.5 Chat wins 7, and 0 benchmarks are tied. The table shows the modelled score for each model and the gap.
| Benchmark | DeepSeek-V3.2 Base | Kimi K1.5 Chat | Gap | Winner |
|---|---|---|---|---|
| MMLU-Pro | 84.3 | 81.3 | 3.0 | DeepSeek-V3.2 Base |
| GPQA Diamond | 77.1 | 78.1 | 1.0 | Kimi K1.5 Chat |
| SWE-bench Verified | 69.1 | 69.9 | 0.8 | Kimi K1.5 Chat |
| HumanEval | 82.8 | 84.3 | 1.5 | Kimi K1.5 Chat |
| MATH-500 | 83.7 | 81.7 | 2.0 | DeepSeek-V3.2 Base |
| AIME 2025 | 74.3 | 81.4 | 7.1 | Kimi K1.5 Chat |
| LiveCodeBench | 73.1 | 75.7 | 2.6 | Kimi K1.5 Chat |
| IFEval | 86.9 | 88.2 | 1.3 | Kimi K1.5 Chat |
| TruthfulQA | 83.5 | 80.3 | 3.2 | DeepSeek-V3.2 Base |
| HellaSwag | 90.2 | 87.2 | 3.0 | DeepSeek-V3.2 Base |
| MMMU | 76.7 | 73.6 | 3.1 | DeepSeek-V3.2 Base |
| Arena Elo | 1417.9 | 1422.6 | 4.7 | Kimi K1.5 Chat |
Price analysis
On a blended 3:1 input-to-output basis, DeepSeek-V3.2 Base costs about $10.56 per million tokens and Kimi K1.5 Chat costs about $10.48. Kimi K1.5 Chat is roughly 1% cheaper than DeepSeek-V3.2 Base. At high volume, that gap compounds quickly, so price should carry real weight unless the quality difference is decisive.
Both models list a 128,000-token context window. That is enough for long documents, large codebases and multi-step agent traces.
Verdict
Across 12 shared benchmarks, DeepSeek-V3.2 Base wins 5, Kimi K1.5 Chat wins 7, and 0 are tied. The two models are effectively tied on the index average at 80.2.
The price picture favours Kimi K1.5 Chat. The context picture is effectively a tie. Those two dimensions often matter more than a few benchmark points once a model is good enough for the task.
If you need a single default, start with Kimi K1.5 Chat if cost is the constraint and DeepSeek-V3.2 Base if context is the constraint. Test both on a sample of your own traffic before committing.
Choose DeepSeek-V3.2 Base if
- You need the larger 128,000-token output limit.
Choose Kimi K1.5 Chat if
- Unit cost matters and you want the cheaper blended price.
FAQ
Is DeepSeek-V3.2 Base better than Kimi K1.5 Chat?
The two models are effectively tied on the index average at 80.2. The better choice depends on price, context and your task. DeepSeek-V3.2 Base wins 5 benchmarks, Kimi K1.5 Chat wins 7.
Which is cheaper, DeepSeek-V3.2 Base or Kimi K1.5 Chat?
Kimi K1.5 Chat is cheaper on a blended 3:1 basis at about $10.48 per million tokens, versus $10.56 for DeepSeek-V3.2 Base.
Which has the larger context window?
Both models list a 128,000-token context window.
Are these models open weights?
DeepSeek-V3.2 Base is open weights; Kimi K1.5 Chat is open weights.
Which should I choose for production?
If you need a single default, start with Kimi K1.5 Chat if cost is the constraint and DeepSeek-V3.2 Base if context is the constraint. Test both on a sample of your own traffic before committing.