AmazonNova 2.0 Lite (high)
Best configuration ranks #282 of 610 on the BenchLeader Index at 50.6 ±7.7 (thinking reasoning effort). Released 29 Oct 2025.
- Blended price
- $0.850/M
- $0.300 in · $2.50 out
- Output speed
- 151 tok/s
- First answer
- 30 s
- first token 1.00 s
- Context
- 1M
- Overall index51
- Agents & tools48
- Knowledge38
- Instruction following69
- Multimodal48
- Long context56
- Composite46
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Default” means the publisher did not say which setting was used.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| low | 47.2 | #353 | 162 tok/s | 24 s | $0.0009 | Agents & tools 37 · Composite 44 · Instruction following 61 · Knowledge 38 · Long context 53 · Multimodal 42 |
| medium | 50.0 | #299 | 151 tok/s | 28 s | $0.0009 | Agents & tools 49 · Composite 45 · Instruction following 67 · Knowledge 37 · Long context 56 · Multimodal 47 |
| thinkingbest | 50.6 | #282 | 151 tok/s | 30 s | $0.0009 | Agents & tools 48 · Composite 46 · Instruction following 69 · Knowledge 38 · Long context 56 · Multimodal 48 |
| default | 41.7 | #479 | 158 tok/s | 1.00 s | $0.0009 | Agents & tools 40 · Composite 40 · Instruction following 43 · Knowledge 36 · Long context 34 · Multimodal 33 |
Benchmark results
One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | low | medium | thinking | default | Source |
|---|---|---|---|---|---|
| GPQA Diamond (AA)not in index | 69.8%#289 | 76.8%#220 | 81.1%#178 | 60.3%#355 | Artificial Analysis |
| Humanity's Last Exam (AA)not in index | 4.0%#463 | 9.0%#271 | 11.6%#227 | 2.9%#517 | Artificial Analysis |
Agents & tools
| Benchmark | low | medium | thinking | default | Source |
|---|---|---|---|---|---|
| Terminal-Bench Hard | 3.8%#283 | 17.4%#173 | 16.7%#179 | 6.8%#240 | Artificial Analysis |
| τ²-Bench Telecom (AA)not in index | 71.9%#145 | 75.7%#132 | 72.8%#142 | 62.0%#172 | Artificial Analysis |
Knowledge
| Benchmark | low | medium | thinking | default | Source |
|---|---|---|---|---|---|
| AA-Omniscience | -54.1#351 | -57.0#366 | -55.0#361 | -59.8#388 | Artificial Analysis |
Instruction following
| Benchmark | low | medium | thinking | default | Source |
|---|---|---|---|---|---|
| IFBench | 61.2%#119 | 68.5%#79 | 70.8%#64 | 40.5%#247 | Artificial Analysis |
Multimodal
| Benchmark | low | medium | thinking | default | Source |
|---|---|---|---|---|---|
| MMMU-Pro | 58.0%#187 | 62.5%#161 | 63.8%#155 | 49.0%#208 | Artificial Analysis |
Long context
| Benchmark | low | medium | thinking | default | Source |
|---|---|---|---|---|---|
| AA-LCR | 54.3%#238 | 60.0%#219 | 60.3%#216 | 18.7%#369 | Artificial Analysis |
Composite
| Benchmark | low | medium | thinking | default | Source |
|---|---|---|---|---|---|
| AA Intelligence Index | 11.8#278 | 12.5#261 | 13.4#245 | 8.7#353 | Artificial Analysis |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0009 | – | 32.2 s |
| Summarise a 30-page report | 12,000 / 600 | $0.0051 | – | 34.2 s |
| Code edit | 6,000 / 1,500 | $0.0056 | – | 40.1 s |
| Agentic coding session | 60,000 / 4,000 | $0.028 | – | 56.6 s |
| Structured extraction | 2,000 / 200 | $0.0011 | – | 31.5 s |
See also
Data as of 9 Sept 2026. Compare these configurations.