Provider Snapshot
15
60.92
718.67
Feb 9, 2026
Key Takeaways
15 bedrock models are actively benchmarked with 21152 total measurements across 18503 benchmark runs.
nova-micro leads the fleet with 127.00 tokens/second, while nova-pro delivers 83.00 tok/s.
Performance varies by 53.0% across the bedrock model lineup, indicating diverse optimization strategies for different use cases.
Avg time to first token across the fleet is 718.67 ms, showing good responsiveness for interactive applications.
The bedrock model fleet shows varied performance characteristics (60.9% variation coefficient), reflecting diverse model architectures.
Fastest Models
| Provider | Model | Avg Toks/Sec | Min | Max | Avg TTF (ms) |
|---|---|---|---|---|---|
| bedrock | nova-micro | 127.00 | 69.20 | 154.00 | 260.00 |
| bedrock | llama-4-maverick | 109.00 | 28.70 | 149.00 | 250.00 |
| bedrock | nova-lite | 102.00 | 42.70 | 135.00 | 290.00 |
| bedrock | llama-4-scout | 102.00 | 1.14 | 140.00 | 340.00 |
| bedrock | llama-3.3-70b | 101.00 | 15.10 | 137.00 | 230.00 |
| bedrock | nova-pro | 83.00 | 10.60 | 124.00 | 410.00 |
All Models
Complete list of all bedrock models tracked in the benchmark system. Click any model name to view detailed performance data.
| Provider | Model | Avg Toks/Sec | Min | Max | Avg TTF (ms) |
|---|---|---|---|---|---|
| bedrock | claude-3-5-haiku | 28.90 | 1.02 | 37.70 | 1270.00 |
| bedrock | claude-3-5-sonnet | 32.60 | 7.26 | 42.00 | 570.00 |
| bedrock | claude-3-7-sonnet | 33.00 | 1.03 | 41.30 | 790.00 |
| bedrock | claude-3-opus | 19.00 | 4.61 | 21.80 | 840.00 |
| bedrock | claude-haiku-4.5 | 45.20 | 8.37 | 64.50 | 910.00 |
| bedrock | claude-sonnet-4.5 | 21.60 | 1.12 | 28.80 | 1810.00 |
| bedrock | nova-lite | 102.00 | 42.70 | 135.00 | 290.00 |
| bedrock | nova-micro | 127.00 | 69.20 | 154.00 | 260.00 |
| bedrock | nova-pro | 83.00 | 10.60 | 124.00 | 410.00 |
| bedrock | llama-3.2-90b | 47.50 | 29.60 | 50.30 | 340.00 |
| bedrock | llama-3.3-70b | 101.00 | 15.10 | 137.00 | 230.00 |
| bedrock | llama-4-maverick | 109.00 | 28.70 | 149.00 | 250.00 |
| bedrock | llama-4-scout | 102.00 | 1.14 | 140.00 | 340.00 |
| bedrock | mistral-large | 44.00 | 8.32 | 47.00 | 240.00 |
| bedrock | claude-opus-4.5 | 18.00 | 1.03 | 22.60 | 2230.00 |
Featured Models
Frequently Asked Questions
Based on recent tests, nova-micro shows the highest average throughput among tracked bedrock models.
This provider summary aggregates 21152 individual prompts measured across 18503 monitoring runs over the past month.