Amazon AWS Bedrockproprietary
Released: 2024-12-03Amazon Nova Micro 1.0
Text-only ultra-low latency model delivering sub-second response times at an industry-disruptive price point.
Intelligence Index
73.0/ 100
Calibrated multi-domain compositeThroughput Speed
240tok/sec
Output streaming throughputTime To First Token
140ms
Initial chunk server latencyLiveBench Contam-Free
64.2%
September 2026 suite scoreSourced Benchmark Evaluations (61)
Verified performance across authoritative benchmarks with provenance tracking.| Benchmark | Status | Score | Trust | Date | Source Type | Provenance |
|---|---|---|---|---|---|---|
| LiveBench (Continuous) | Active | 64.2% | 94 | 2025-02-01 | independent | Source ↗ |
| ARC-AGI-3 | Active | 46.2% | 93 | 2025-01-07 | independent | Source ↗ |
| ARC-AGI-2 | Nearing Saturation | 55.4% | 75 | 2024-12-31 | independent | Source ↗ |
| FrontierMath | Active | 19.3% | 98 | 2024-12-28 | independent | Source ↗ |
| PutnamBench | Active | 35.9% | 90 | 2024-12-27 | independent | Source ↗ |
| Omni-MATH | Deprecated | 51.1% | 10 | 2024-12-25 | independent | Source ↗ |
| SWE-Lancer | Deprecated | 43.2% | 20 | 2024-12-25 | independent | Source ↗ |
| Soohak | Active | 52.4% | 98 | 2024-12-23 | independent | Source ↗ |
| FrontierCode | Active | 36.8% | 98 | 2024-12-23 | independent | Source ↗ |
| Humanity's Last Exam | Active | 31.2% | 93 | 2024-12-21 | independent | Source ↗ |
| StructFlowBench | Nearing Saturation | 57.1% | 75 | 2024-12-21 | independent | Source ↗ |
| AIME | Nearing Saturation | 65.6% | 75 | 2024-12-21 | independent | Source ↗ |
| OlympiadBench | Active | 56.6% | 80 | 2024-12-20 | independent | Source ↗ |
| FireBench | Active | 56.4% | 95 | 2024-12-19 | independent | Source ↗ |
| Terminal-Bench | Active | 44.4% | 80 | 2024-12-19 | independent | Source ↗ |
| MathArena | Active | 55.2% | 99 | 2024-12-19 | independent | Source ↗ |
| OpenAI-MRCR | Active | 61.3% | 80 | 2024-12-18 | independent | Source ↗ |
| CodeContests | Deprecated | 41.3% | 10 | 2024-12-18 | independent | Source ↗ |
| SimpleQA | Active | 38% | 90 | 2024-12-18 | independent | Source ↗ |
| ComplexBench | Nearing Saturation | 55.1% | 80 | 2024-12-17 | independent | Source ↗ |
| SWE-bench Pro | Active | 39.4% | 90 | 2024-12-17 | independent | Source ↗ |
| LongBench-Pro | Active | 59.9% | 80 | 2024-12-17 | independent | Source ↗ |
| MultiPL-E | Deprecated | 58.4% | 10 | 2024-12-16 | independent | Source ↗ |
| ZeroSCROLLS | Nearing Saturation | 58.4% | 78 | 2024-12-16 | independent | Source ↗ |
| AgentIF | Nearing Saturation | 45.7% | 80 | 2024-12-15 | independent | Source ↗ |
| MATH | Saturated | 67.6% | 25 | 2024-12-15 | independent | Source ↗ |
| LiveCodeBench | Saturated | 52.1% | 39 | 2024-12-15 | independent | Source ↗ |
| InfiniteBench | Nearing Saturation | 62.1% | 65 | 2024-12-15 | independent | Source ↗ |
| MMLU-Pro | Nearing Saturation | 67.2% | 75 | 2024-12-14 | independent | Source ↗ |
| Mercury | Deprecated | 55.9% | 10 | 2024-12-14 | independent | Source ↗ |
| SCROLLS | Nearing Saturation | 58.4% | 65 | 2024-12-14 | independent | Source ↗ |
| RULER | Nearing Saturation | 67.2% | 84 | 2024-12-13 | independent | Source ↗ |
| TabMWP | Saturated | 63.5% | 25 | 2024-12-13 | independent | Source ↗ |
| CyberSecEval | Deprecated | 57.2% | 10 | 2024-12-13 | independent | Source ↗ |
| MMLU-Redux | Saturated | 71.5% | 25 | 2024-12-12 | independent | Source ↗ |
| QuALITY | Nearing Saturation | 64.2% | 78 | 2024-12-12 | independent | Source ↗ |
| SWE-bench Verified | Saturated | 49.5% | 25 | 2024-12-12 | independent | Source ↗ |
| MGSM | Nearing Saturation | 65.6% | 65 | 2024-12-11 | independent | Source ↗ |
| LongBench | Nearing Saturation | 62.8% | 65 | 2024-12-11 | independent | Source ↗ |
| SWE-bench | Deprecated | 43.2% | 10 | 2024-12-11 | independent | Source ↗ |
| AGIEval | Saturated | 61.1% | 25 | 2024-12-10 | independent | Source ↗ |
| QASPER | Nearing Saturation | 60.6% | 65 | 2024-12-10 | independent | Source ↗ |
| TruthfulQA | Saturated | 64.2% | 25 | 2024-12-10 | independent | Source ↗ |
| SuperGLUE | Saturated | 72.4% | 25 | 2024-12-09 | independent | Source ↗ |
| GSM-Hard | Deprecated | 66.2% | 10 | 2024-12-09 | independent | Source ↗ |
| MT-Bench | Nearing Saturation | 6.9% | 75 | 2024-12-08 | independent | Source ↗ |
| BIG-Bench Hard | Saturated | 69.6% | 25 | 2024-12-08 | independent | Source ↗ |
| CodeXGLUE | Deprecated | 59.7% | 23 | 2024-12-08 | independent | Source ↗ |
| MBPP+ | Nearing Saturation | 64.8% | 65 | 2024-12-07 | independent | Source ↗ |
| WinoGrande | Saturated | 73.8% | 28 | 2024-12-07 | independent | Source ↗ |
| IFEval | Deprecated | 65.8% | 25 | 2024-12-07 | independent | Source ↗ |
| GPQA | Nearing Saturation | 64.2% | 78 | 2024-12-07 | independent | Source ↗ |
| HumanEval+ | Saturated | 68.6% | 25 | 2024-12-06 | independent | Source ↗ |
| Needle-in-a-Haystack | Saturated | 94.6% | 44 | 2024-12-06 | independent | Source ↗ |
| CommonsenseQA | Saturated | 74.6% | 28 | 2024-12-06 | independent | Source ↗ |
| GSM8K | Saturated | 74.5% | 25 | 2024-12-05 | independent | Source ↗ |
| GLUE | Saturated | 76.7% | 28 | 2024-12-05 | independent | Source ↗ |
| MBPP | Saturated | 67.3% | 25 | 2024-12-05 | independent | Source ↗ |
| HumanEval | Saturated | 72.4% | 25 | 2024-12-04 | independent | Source ↗ |
| MMLU | Saturated | 74.5% | 25 | 2024-12-04 | independent | Source ↗ |
| ARC-AGI | Saturated | 66.7% | 25 | 2024-12-04 | independent | Source ↗ |
OpenAI Compatible API ExecutionAPI Ready
from openai import OpenAI
client = OpenAI()
response = client.chat.completions.create(
model="amazon-nova-micro-v1",
messages=[{"role": "user", "content": "Evaluate multi-step logic problem"}],
temperature=0.2
)
print(response.choices[0].message.content)Evaluation & Sourcing Notes
Scores listed for Amazon Nova Micro 1.0 represent verified evaluations extracted from official research papers, independent evaluation suites (HELM, LMSYS, OpenCompass, LiveBench), and verified audit reports.
All benchmarks marked as saturated or deprecated reflect historical performance where the benchmark no longer provides active discriminative power.