Compare Models
Select up to 4 models to compare benchmarks, pricing, and capabilities side by side.
OpenAI
Moonshot AI
Add Model
MMLU
GPT-4o
88.7
Gemini 2.5 Pro
90.2
Kimi K2.5
92.0
HumanEval
GPT-4o
90.2
Gemini 2.5 Pro
93.0
Kimi K2.5
99.0
GSM8K
GPT-4o
95.8
Gemini 2.5 Pro
97.5
Kimi K2.5
99.0
GPQA
GPT-4o
53.6
Gemini 2.5 Pro
72.0
Kimi K2.5
87.6
MGSM
GPT-4o
90.5
Gemini 2.5 Pro
94.0
Kimi K2.5
96.0
ARC-Challenge
GPT-4o
96.7
Gemini 2.5 Pro
97.0
Kimi K2.5
0.0
HellaSwag
GPT-4o
95.3
Gemini 2.5 Pro
95.5
Kimi K2.5
0.0
MATH
GPT-4o
76.6
Gemini 2.5 Pro
91.8
Kimi K2.5
98.0
SWE-bench
GPT-4o
38.4
Gemini 2.5 Pro
63.8
Kimi K2.5
76.8
MMMLU
GPT-4o
85.1
Gemini 2.5 Pro
89.5
Kimi K2.5
0.0
LiveCodeBench
GPT-4o
0.0
Gemini 2.5 Pro
0.0
Kimi K2.5
85.0
IFEval
GPT-4o
0.0
Gemini 2.5 Pro
0.0
Kimi K2.5
94.0
AIME 2025
GPT-4o
0.0
Gemini 2.5 Pro
0.0
Kimi K2.5
96.1
| Model | Input | Output | Blended* |
|---|---|---|---|
GPT-4o | $2.50 | $10.00 | $6.25 |
Gemini 2.5 Pro | $1.25 | $10.00 | $5.63 |
Kimi K2.5 | $0.45 | $2.20 | $1.33 |
*Blended = average of input and output price
| Spec | GPT-4o | Gemini 2.5 Pro | Kimi K2.5 |
|---|---|---|---|
| Context Window | 128K | 1.0M | 256K |
| Max Output | 16K | 66K | 16K |
| TTFT | 320ms | 600ms | 500ms |
| Speed | 95 tok/s | 85 tok/s | 70 tok/s |
| Parameters | ~1.8T (estimated) | N/A | 1T (32B active) |
| Architecture | Transformer (MoE) | Transformer (MoE) + Thinking | MoE + Multimodal |
| Open Source | No | No | No |
| Tier | frontier | frontier | frontier |
Quick Verdict
Best Performance
Kimi K2.5
Best Value
Kimi K2.5
Fastest
GPT-4o