kimi-k3 vs gemini-3.8-flash
Side-by-side comparison of kimi-k3 and gemini-3.8-flash: benchmarks, pricing, context window and capabilities. Both are accessible through Requesty's unified API. gemini-3.8-flash outperforms kimi-k3 on 3 of 5 shared benchmarks.
kimi-k3
Input / 1M
$3.00
Output / 1M
$15.00
Context
1.0M
Model ID
moonshot/kimi-k3
gemini-3.8-flash
Input / 1M
$1.50$0.75
Output / 1M
$7.50$3.75
Context
1.0M
Model ID
vertex/gemini-3.8-flash
Benchmark comparison
Intelligence Indexreasoning
kimi-k350.2%
gemini-3.8-flash47.1%
Coding Indexcoding
kimi-k376.2%
gemini-3.8-flash76.3%
GPQA Diamondreasoning
kimi-k393.5%
gemini-3.8-flash95.3%
SciCodecoding
kimi-k359.5%
gemini-3.8-flash56.6%
Humanity's Last Examreasoning
kimi-k346.9%
gemini-3.8-flash47.8%
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
| kimi-k3 | gemini-3.8-flash | |
|---|---|---|
| Input price / 1M | $3.00 | $1.50$0.75 |
| Output price / 1M | $15.00 | $7.50$3.75 |
| Context window | 1.0M tokens | 1.0M tokens |
| Max output | 1.0M tokens | 66K tokens |
| Vision input | Yes | Yes |
| Tool calling | Yes | Yes |
| Reasoning | Yes | Yes |
| Prompt caching | Yes | Yes |
| Computer use | N/A | N/A |
| Provider | Moonshot AI | Google LLC (Vertex AI) |
Questions people ask
Is kimi-k3 better than gemini-3.8-flash?
gemini-3.8-flash outperforms kimi-k3 on 3 of 5 shared benchmarks. See the benchmark comparison above for specifics: kimi-k3 and gemini-3.8-flash have different strengths across reasoning, coding, math and multimodal tasks.
Which is cheaper, kimi-k3 or gemini-3.8-flash?
gemini-3.8-flash is cheaper. kimi-k3 costs $3.00/$15.00 per 1M input/output tokens, while gemini-3.8-flash costs $0.75/$3.75. The prices shown are what you pay: gemini-3.8-flash runs at 50% off its list rate of $1.50 per 1M input and $7.50 per 1M output.
Can I use kimi-k3 and gemini-3.8-flash through the same API?
Yes. Requesty provides a single OpenAI-compatible API that routes to both. Change just the "model" parameter to switch between "moonshot/kimi-k3" and "vertex/gemini-3.8-flash", no other code changes needed.
What are the context windows?
kimi-k3 supports up to 1.0M tokens of context. gemini-3.8-flash supports up to 1.0M 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 kimi-k3 and gemini-3.8-flash with one line of code
Requesty provides a single OpenAI-compatible API for 600+ models. Change the model parameter, not your code.