All comparisons
HellaSwag — 96.4 HumanEval — 90.2 Massive Multitask Language Understanding — 88.7
Kimi K2.5 vs GPT-4o (05-13)
Side-by-side comparison of specifications, capabilities, and pricing.
| Released | 2026-03-15 | 2024-05-13 |
| Context window | 256K | 128K |
| Parameters | 1T (MoE, 384 experts) | 1.76T (8x222B MoE)* |
| Architecture | mixture of experts | mixture of experts |
| License | Unknown | Proprietary |
| Knowledge cutoff | — | 2023-10 |
Capabilities | ||
| Vision | ||
| Multimodal | ||
| Reasoning | ||
| Function calling | ||
| Tool use | ||
| JSON mode | ||
| Code execution | ||
Availability | ||
| Providers | ||
Benchmarks