Amazon AWS Bedrockproprietary
Released: 2025-03-10Amazon Nova Premier 1.0
Amazon most capable multimodal foundation model engineered for advanced enterprise agentic workflows, long-document analytics, and code generation.
Intelligence Index
90.8/ 100
Calibrated multi-domain compositeThroughput Speed
72tok/sec
Output streaming throughputTime To First Token
390ms
Initial chunk server latencyLiveBench Contam-Free
78.2%
September 2026 suite scoreSourced Benchmark Evaluations (61)
Verified performance across authoritative benchmarks with provenance tracking.| Benchmark | Status | Score | Trust | Date | Source Type | Provenance |
|---|---|---|---|---|---|---|
| ARC-AGI-3 | Active | 58.8% | 93 | 2025-04-14 | independent | Source ↗ |
| ARC-AGI-2 | Nearing Saturation | 70.6% | 75 | 2025-04-07 | independent | Source ↗ |
| FrontierMath | Active | 24.9% | 98 | 2025-04-04 | independent | Source ↗ |
| PutnamBench | Active | 46.3% | 90 | 2025-04-03 | independent | Source ↗ |
| SWE-Lancer | Deprecated | 55.4% | 20 | 2025-04-01 | independent | Source ↗ |
| Omni-MATH | Deprecated | 65.9% | 10 | 2025-04-01 | independent | Source ↗ |
| Soohak | Active | 67.6% | 98 | 2025-03-30 | independent | Source ↗ |
| FrontierCode | Active | 47.3% | 98 | 2025-03-30 | independent | Source ↗ |
| Humanity's Last Exam | Active | 39.8% | 93 | 2025-03-28 | independent | Source ↗ |
| StructFlowBench | Nearing Saturation | 71% | 75 | 2025-03-28 | independent | Source ↗ |
| AIME | Nearing Saturation | 84.6% | 75 | 2025-03-28 | independent | Source ↗ |
| OlympiadBench | Active | 73% | 80 | 2025-03-27 | independent | Source ↗ |
| MathArena | Active | 71.2% | 99 | 2025-03-26 | independent | Source ↗ |
| FireBench | Active | 70.2% | 95 | 2025-03-26 | independent | Source ↗ |
| Terminal-Bench | Active | 57.1% | 80 | 2025-03-26 | independent | Source ↗ |
| OpenAI-MRCR | Active | 76.3% | 80 | 2025-03-25 | independent | Source ↗ |
| SimpleQA | Active | 47.2% | 90 | 2025-03-25 | independent | Source ↗ |
| CodeContests | Deprecated | 53% | 10 | 2025-03-25 | independent | Source ↗ |
| ComplexBench | Nearing Saturation | 68.5% | 80 | 2025-03-24 | independent | Source ↗ |
| SWE-bench Pro | Active | 50.5% | 90 | 2025-03-24 | independent | Source ↗ |
| LongBench-Pro | Active | 74.5% | 80 | 2025-03-24 | independent | Source ↗ |
| ZeroSCROLLS | Nearing Saturation | 72.6% | 78 | 2025-03-23 | independent | Source ↗ |
| MultiPL-E | Deprecated | 75% | 10 | 2025-03-23 | independent | Source ↗ |
| LiveCodeBench | Saturated | 66.8% | 39 | 2025-03-22 | independent | Source ↗ |
| AgentIF | Nearing Saturation | 56.8% | 80 | 2025-03-22 | independent | Source ↗ |
| MATH | Saturated | 87.2% | 25 | 2025-03-22 | independent | Source ↗ |
| InfiniteBench | Nearing Saturation | 77.2% | 65 | 2025-03-22 | independent | Source ↗ |
| Mercury | Deprecated | 71.7% | 10 | 2025-03-21 | independent | Source ↗ |
| SCROLLS | Nearing Saturation | 72.6% | 65 | 2025-03-21 | independent | Source ↗ |
| MMLU-Pro | Nearing Saturation | 83.5% | 75 | 2025-03-21 | independent | Source ↗ |
| RULER | Nearing Saturation | 83.5% | 84 | 2025-03-20 | independent | Source ↗ |
| TabMWP | Saturated | 81.9% | 25 | 2025-03-20 | independent | Source ↗ |
| CyberSecEval | Deprecated | 73.4% | 10 | 2025-03-20 | independent | Source ↗ |
| SWE-bench Verified | Saturated | 63.6% | 25 | 2025-03-19 | independent | Source ↗ |
| MMLU-Redux | Saturated | 89% | 25 | 2025-03-19 | independent | Source ↗ |
| QuALITY | Nearing Saturation | 79.9% | 78 | 2025-03-19 | independent | Source ↗ |
| LongBench | Nearing Saturation | 78.1% | 65 | 2025-03-18 | independent | Source ↗ |
| SWE-bench | Deprecated | 55.4% | 10 | 2025-03-18 | independent | Source ↗ |
| MGSM | Nearing Saturation | 84.6% | 65 | 2025-03-18 | independent | Source ↗ |
| AGIEval | Saturated | 77.8% | 25 | 2025-03-17 | independent | Source ↗ |
| QASPER | Nearing Saturation | 75.4% | 65 | 2025-03-17 | independent | Source ↗ |
| TruthfulQA | Saturated | 79.9% | 25 | 2025-03-17 | independent | Source ↗ |
| GSM-Hard | Deprecated | 85.4% | 10 | 2025-03-16 | independent | Source ↗ |
| SuperGLUE | Saturated | 92.3% | 25 | 2025-03-16 | independent | Source ↗ |
| CodeXGLUE | Deprecated | 76.6% | 23 | 2025-03-15 | independent | Source ↗ |
| MT-Bench | Nearing Saturation | 8.6% | 75 | 2025-03-15 | independent | Source ↗ |
| BIG-Bench Hard | Saturated | 88.7% | 25 | 2025-03-15 | independent | Source ↗ |
| WinoGrande | Saturated | 94.1% | 28 | 2025-03-14 | independent | Source ↗ |
| MBPP+ | Nearing Saturation | 83.1% | 65 | 2025-03-14 | independent | Source ↗ |
| IFEval | Deprecated | 81.9% | 25 | 2025-03-14 | independent | Source ↗ |
| GPQA | Nearing Saturation | 79.9% | 78 | 2025-03-14 | independent | Source ↗ |
| Needle-in-a-Haystack | Saturated | 98.2% | 44 | 2025-03-13 | independent | Source ↗ |
| HumanEval+ | Saturated | 88% | 25 | 2025-03-13 | independent | Source ↗ |
| CommonsenseQA | Saturated | 95% | 28 | 2025-03-13 | independent | Source ↗ |
| MBPP | Saturated | 86.4% | 25 | 2025-03-12 | independent | Source ↗ |
| GSM8K | Saturated | 96.1% | 25 | 2025-03-12 | independent | Source ↗ |
| GLUE | Saturated | 97.7% | 28 | 2025-03-12 | independent | Source ↗ |
| MMLU | Saturated | 92.6% | 25 | 2025-03-11 | independent | Source ↗ |
| HumanEval | Saturated | 92.9% | 25 | 2025-03-11 | independent | Source ↗ |
| ARC-AGI | Saturated | 85.1% | 25 | 2025-03-11 | independent | Source ↗ |
| LiveBench (Continuous) | Active | 78.2% | 94 | 2025-02-01 | independent | Source ↗ |
OpenAI Compatible API ExecutionAPI Ready
from openai import OpenAI
client = OpenAI()
response = client.chat.completions.create(
model="amazon-nova-premier-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 Premier 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.