Updated March 26, 2026· Based on independent benchmark data
Gemini 3.1 Pro Preview leads in intelligence with a score of 57.2 vs 9.7. Llama 3.2 Instruct 3B is 23.5x cheaper at $0.08/1M tokens vs $2.00/1M. For speed, Gemini 3.1 Pro Preview wins at 113 tok/s vs 51 tok/s.
| Metric | Llama 3.2 Instruct 3B | Gemini 3.1 Pro Preview |
|---|---|---|
| Intelligence Score | 9.7 | 57.2 |
| Coding Score | N/A | 55.5 |
| Math Score | 3.3 | N/A |
| Speed (tok/s) | 51 tok/s | 113 tok/s |
| Latency (TTFT) | 0.38s | 23.84s |
| Input Price / 1M tokens | $0.08 | $2.00 |
| Output Price / 1M tokens | $0.08 | $12 |
| Context Window | N/A | 1.0M |
Gemini 3.1 Pro Preview outperforms Llama 3.2 Instruct 3B on the intelligence index with a score of 57.2 compared to 9.7.
Gemini 3.1 Pro Preview generates output significantly faster at 113 tok/s compared to Llama 3.2 Instruct 3B's 51 tok/s, making it 2.2x faster for streaming responses. Time to first token is 0.38s for Llama 3.2 Instruct 3B vs 23.84s for Gemini 3.1 Pro Preview, which affects perceived responsiveness in interactive applications.
Llama 3.2 Instruct 3B is more affordable at $0.08/1M input tokens ($0.08/1M output), while Gemini 3.1 Pro Preview costs $2.00/1M input ($12/1M output). That makes Gemini 3.1 Pro Preview 23.5x more expensive per token, which can add up significantly at scale. For a typical workload of 100 requests per day at 2,000 tokens each, Llama 3.2 Instruct 3B would cost approximately $0.51/month vs $12.00/month for Gemini 3.1 Pro Preview in input costs alone.
Choose Llama 3.2 Instruct 3B when you need lower cost. Choose Gemini 3.1 Pro Preview when you need higher intelligence (57.2), faster output (113 tok/s).
Llama 3.2 Instruct 3B is cheaper at $0.08/1M input tokens vs $2.00/1M for Gemini 3.1 Pro Preview.
Gemini 3.1 Pro Preview is faster, producing output at 113 tok/s compared to Llama 3.2 Instruct 3B's 51 tok/s.
No, Llama 3.2 Instruct 3B does not support image input. However, Gemini 3.1 Pro Preview does support images.
It depends on your priorities. Gemini 3.1 Pro Preview scores higher on intelligence (57.2), but Llama 3.2 Instruct 3B may be better for specific use cases like budget-conscious projects or speed-critical applications.
Data last synced: March 26, 2026
| Max Output Tokens | N/A | N/A |
| Input Modalities | Text | Audio + File + Image + Text + Video |
| Output Modalities | Text | Text |
| Free Tier | No | No |