deepseek-v4-flash vs kimi-k2
Side-by-side comparison of deepseek-v4-flash and kimi-k2: benchmarks, pricing, context window and capabilities. Both are accessible through Requesty's unified API. deepseek-v4-flash outperforms kimi-k2 on 7 of 7 shared benchmarks.
deepseek-v4-flash
Input / 1M
$0.14
Output / 1M
$0.28
Context
1M
Model ID
deepseek/deepseek-v4-flash
kimi-k2
Input / 1M
$0.60
Output / 1M
$2.50
Context
262K
Model ID
vertex/kimi-k2
Benchmark comparison
Intelligence Indexreasoning
deepseek-v4-flash46.5%
kimi-k240.9%
Coding Indexcoding
deepseek-v4-flash38.7%
kimi-k234.8%
Math Indexmath
deepseek-v4-flashN/A
kimi-k294.7%
GPQA Diamondreasoning
deepseek-v4-flash89.4%
kimi-k283.8%
AIME 2025math
deepseek-v4-flashN/A
kimi-k294.7%
LiveCodeBenchcoding
deepseek-v4-flashN/A
kimi-k285.3%
Terminal-Bench Hardagentic
deepseek-v4-flash35.6%
kimi-k231.1%
τ²-Benchagentic
deepseek-v4-flash95.0%
kimi-k293.0%
SciCodecoding
deepseek-v4-flash44.9%
kimi-k242.4%
MMLU Proknowledge
deepseek-v4-flashN/A
kimi-k284.8%
Humanity's Last Examreasoning
deepseek-v4-flash32.1%
kimi-k222.3%
Scores sourced from official model cards, Artificial Analysis, and public leaderboards. Benchmarks measure specific skills and don't capture every aspect of model quality.
Pricing & specifications
| deepseek-v4-flash | kimi-k2 | |
|---|---|---|
| Input price / 1M | $0.14 | $0.60 |
| Output price / 1M | $0.28 | $2.50 |
| Context window | 1M tokens | 262K tokens |
| Max output | 384K tokens | 262K tokens |
| Vision input | N/A | Yes |
| Tool calling | Yes | Yes |
| Reasoning | N/A | Yes |
| Prompt caching | Yes | Yes |
| Computer use | N/A | N/A |
| Provider | DeepSeek | Google LLC (Vertex AI) |
Questions people ask
Is deepseek-v4-flash better than kimi-k2?
deepseek-v4-flash outperforms kimi-k2 on 7 of 7 shared benchmarks. See the benchmark comparison above for specifics: deepseek-v4-flash and kimi-k2 have different strengths across reasoning, coding, math and multimodal tasks.
Which is cheaper, deepseek-v4-flash or kimi-k2?
deepseek-v4-flash is cheaper. deepseek-v4-flash costs $0.14/$0.28 per 1M input/output tokens, while kimi-k2 costs $0.60/$2.50.
Can I use deepseek-v4-flash and kimi-k2 through the same API?
Yes. Requesty provides a single OpenAI-compatible API that routes to both. Change just the "model" parameter to switch between "deepseek/deepseek-v4-flash" and "vertex/kimi-k2", no other code changes needed.
What are the context windows?
deepseek-v4-flash supports up to 1M tokens of context. kimi-k2 supports up to 262K tokens. Longer context means you can feed larger documents or codebases in a single prompt, though quality often degrades past 128K for most models.
Switch between deepseek-v4-flash and kimi-k2 with one line of code
Requesty provides a single OpenAI-compatible API for 400+ models. Change the model parameter, not your code.
Get started free