kimi-k3 vs grok-4.5
Side-by-side comparison of kimi-k3 and grok-4.5: benchmarks, pricing, context window and capabilities. Both are accessible through Requesty's unified API. kimi-k3 outperforms grok-4.5 on 5 of 5 shared benchmarks.
kimi-k3
Input / 1M
$3.00
Output / 1M
$15.00
Context
1.0M
Model ID
moonshot/kimi-k3
xAI Corp.
grok-4.5
Input / 1M
$2.00
Output / 1M
$6.00
Context
500K
Model ID
xai/grok-4.5
Benchmark comparison
Intelligence Indexreasoning
kimi-k350.2%
grok-4.545.5%
Coding Indexcoding
kimi-k376.2%
grok-4.572.4%
GPQA Diamondreasoning
kimi-k393.5%
grok-4.593.1%
SciCodecoding
kimi-k359.5%
grok-4.555.0%
Humanity's Last Examreasoning
kimi-k346.9%
grok-4.542.7%
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 | grok-4.5 | |
|---|---|---|
| Input price / 1M | $3.00 | $2.00 |
| Output price / 1M | $15.00 | $6.00 |
| Context window | 1.0M tokens | 500K tokens |
| Max output | 1.0M tokens | N/A |
| Vision input | Yes | Yes |
| Tool calling | Yes | Yes |
| Reasoning | Yes | Yes |
| Prompt caching | Yes | Yes |
| Computer use | N/A | N/A |
| Provider | Moonshot AI | xAI Corp. |
Questions people ask
Is kimi-k3 better than grok-4.5?
kimi-k3 outperforms grok-4.5 on 5 of 5 shared benchmarks. See the benchmark comparison above for specifics: kimi-k3 and grok-4.5 have different strengths across reasoning, coding, math and multimodal tasks.
Which is cheaper, kimi-k3 or grok-4.5?
grok-4.5 is cheaper. kimi-k3 costs $3.00/$15.00 per 1M input/output tokens, while grok-4.5 costs $2.00/$6.00.
Can I use kimi-k3 and grok-4.5 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 "xai/grok-4.5", no other code changes needed.
What are the context windows?
kimi-k3 supports up to 1.0M tokens of context. grok-4.5 supports up to 500K 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 grok-4.5 with one line of code
Requesty provides a single OpenAI-compatible API for 600+ models. Change the model parameter, not your code.