GLM 4.7 vs Llama 3.2 1B Instruct
Zhipu AI vs Meta — benchmarks, pricing, and capabilities side by side.
- •GLM 4.7 has the higher intelligence index (89 vs 12.1)
- •Llama 3.2 1B Instruct is cheaper ($0.03 vs $0.40 per 1M input)
- •GLM 4.7 is faster
- •GLM 4.7 has a larger context window (203K)
| GLM 4.7 | Llama 3.2 1B Instruct | |
|---|---|---|
| Intelligence index | 89 | 12.1 |
| Developer | Zhipu AI | Meta |
| Type | LLM | LLM |
| Access | Open weights | Open weights |
| Context window | 202,752 tokens | 131,072 tokens |
| Input price | $0.40 / 1M | $0.03 / 1M |
| Output price | $1.75 / 1M | $0.20 / 1M |
| Speed | 98 tok/s | 91 tok/s |
| Released | December 22, 2025 | September 25, 2024 |
| Parameters | — | — |
| Input modalities | Text | Text |
| Output modalities | Text | Text |
Shared benchmarks
GLM 4.7
Llama 3.2 1B Instruct
AIME 2025
95
0
GPQA Diamond
85.9
19.6
Humanity’s Last Exam
25.1
5.3
LiveCodeBench
89.4
1.9
MMLU-Pro
85.6
20