Updated March 26, 2026· Based on independent benchmark data
Mercury 2 leads in intelligence with a score of 32.8 vs 7.0. Granite 3.3 8B (Non-reasoning) is 8.3x cheaper at $0.03/1M tokens vs $0.25/1M. For speed, Mercury 2 wins at 894 tok/s vs 402 tok/s.
| Metric | Mercury 2 | Granite 3.3 8B (Non-reasoning) |
|---|---|---|
| Intelligence Score | 32.8 | 7.0 |
| Coding Score | 30.6 | 3.4 |
| Math Score | N/A | 6.7 |
| Speed (tok/s) | 894 tok/s | 402 tok/s |
| Latency (TTFT) | 3.81s | 9.53s |
| Input Price / 1M tokens | $0.25 | $0.03 |
| Output Price / 1M tokens | $0.75 | $0.25 |
| Context Window | 128K |
Mercury 2 outperforms Granite 3.3 8B (Non-reasoning) on the intelligence index with a score of 32.8 compared to 7.0. For coding tasks, Mercury 2 has the edge with a coding score of 30.6 vs 3.4.
Mercury 2 generates output significantly faster at 894 tok/s compared to Granite 3.3 8B (Non-reasoning)'s 402 tok/s, making it 2.2x faster for streaming responses. Time to first token is 3.81s for Mercury 2 vs 9.53s for Granite 3.3 8B (Non-reasoning), which affects perceived responsiveness in interactive applications.
Granite 3.3 8B (Non-reasoning) is more affordable at $0.03/1M input tokens ($0.25/1M output), while Mercury 2 costs $0.25/1M input ($0.75/1M output). That makes Mercury 2 8.3x 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, Mercury 2 would cost approximately $1.50/month vs $0.18/month for Granite 3.3 8B (Non-reasoning) in input costs alone.
Choose Mercury 2 when you need higher intelligence (32.8), stronger coding performance (30.6), faster output (894 tok/s). Choose Granite 3.3 8B (Non-reasoning) when you need lower cost.
Mercury 2 scores higher on coding benchmarks (30.6 vs 3.4), making it the better choice for programming tasks.
Granite 3.3 8B (Non-reasoning) is cheaper at $0.03/1M input tokens vs $0.25/1M for Mercury 2.
Mercury 2 is faster, producing output at 894 tok/s compared to Granite 3.3 8B (Non-reasoning)'s 402 tok/s.
No, Mercury 2 does not support image input. Neither model supports image input.
Data last synced: March 26, 2026
| N/A |
| Max Output Tokens | N/A | N/A |
| Input Modalities | Text | Text |
| Output Modalities | Text | Text |
| Free Tier | No | No |
It depends on your priorities. Mercury 2 scores higher on intelligence (32.8), but Granite 3.3 8B (Non-reasoning) may be better for specific use cases like budget-conscious projects or speed-critical applications.