llm-benchmarks
githubdrose.io

bedrock Provider Benchmarks


Measured over the last 30 days

Models Tracked
23
Avg Tokens / Second
61.33
Avg Time to First Token (ms)
761.74
Last Updated
Nov 7, 2026

All bedrock models

23 live · one row per model · fastest first · 1 retired lanes not listed · 5 folded behind their current lane
ModelAvg Toks/SecMinMaxAvg TTF (ms)
Nova Micro 1.0117.0040.30165.00290.00
Nova Lite 1.0108.0022.90147.00330.00
Llama 4 Maverick106.0022.70134.00300.00
Llama 4 Scout95.7025.50125.00300.00
Llama 3.1 8B Instruct95.004.54113.00400.00
Llama 3.3 70B Instruct87.305.04128.00310.00
mistral-7b81.106.7287.20200.00
llama-3-8b77.901.0584.10270.00
Nova Pro 1.077.5027.20125.00410.00
mixtral-8x7b73.5013.1081.00220.00
Llama 3.1 70B Instruct61.5018.5074.20420.00
Claude Haiku 4.559.8012.9086.60550.00
mistral-small56.3012.0060.20220.00
Nova 2 Lite49.7039.2058.90880.00
Mistral Large42.3012.4046.20270.00
llama-3-70b38.9012.1041.90280.00
Claude Sonnet 4.638.805.4247.10770.00
Claude Opus 4.735.009.9946.20970.00
Palmyra X534.309.3444.401880.00
Claude Opus 4.623.501.0032.401580.00
Claude Sonnet 4.521.805.4225.601400.00
Claude Opus 4.521.306.5324.701530.00
Claude Opus 4.18.321.2217.603740.00

Frequently Asked Questions

Which bedrock model is fastest?
Based on recent tests, Nova Micro 1.0 shows the highest average throughput among tracked bedrock models.
How many recent measurements feed this dashboard?
This provider summary aggregates 29600 individual prompts measured across 26657 monitoring runs over the past month.