Compare / head-to-head
Amazon Nova 2 LitevsPhi-4 Reasoning
Amazon Nova 2 Lite leads 3 of 3 shared benchmarks. Amazon Nova 2 Lite has the larger context window (1M tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
3 – 0
Amazon Nova 2 Lite leads
Cheaper per token
—
list price, input + output
Larger context
Amazon Nova 2 Lite
1M tokens
Provider uptime (30d)
— · —
Amazon · Microsoft
Amazon Nova 2 Lite
Amazon · released 2025-12-02
textvision
- Context
- 1M
- Max out
- 66K
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 12 · 3 core
Phi-4 Reasoning
Microsoft · released 2025-04-30
text
- Context
- 33K
- Max out
- 33K
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 4 · 4 core
Quality
Benchmark matrix
3 shared · 9 only Amazon Nova 2 Lite · 1 only Phi-4 Reasoning| Benchmark | Amazon Nova 2 Lite | Phi-4 Reasoning | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 3 | ||||
| GPQA Diamond | 79.6% ↗ | 65.8% ↗ | +13.8 pt | Amazon Nova 2 Lite |
| MMLU-Pro | 80.9% ↗ | 74.3% ↗ | +6.6 pt | Amazon Nova 2 Lite |
| AIME 2025 | 91% ↗ | 62.9% ↗ | +28.1 pt | Amazon Nova 2 Lite |
| Only Amazon Nova 2 Lite reports · 9 | ||||
| BFCL v4 | 60.3% ↗ | not reported | — | — |
| IF-Bench | 70.8% ↗ | not reported | — | — |
| LongCodeBench 1M | 84% ↗ | not reported | — | — |
| MCP Atlas | 24.6% ↗ | not reported | — | — |
| MultiChallenge | 76.6% ↗ | not reported | — | — |
| RealKIE-FCC Verified | 62.1% ↗ | not reported | — | — |
| tau2-bench Airline Verified | 64.8% ↗ | not reported | — | — |
| tau2-bench Retail Verified | 76.5% ↗ | not reported | — | — |
| tau2-bench Telecom | 76% ↗ | not reported | — | — |
| Only Phi-4 Reasoning reports · 1 | ||||
| HumanEval+ | not reported | 92.9% ↗ | — | — |
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is Amazon Nova 2 Lite minus Phi-4 Reasoning in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | Amazon Nova 2 Lite | Phi-4 Reasoning | Edge |
|---|---|---|---|
| Context window | 1M tokens | 33K tokens | Amazon Nova 2 Lite |
| Max output | 66K tokens | 33K tokens | Amazon Nova 2 Lite |
| Input price / 1M | — | — | — |
| Output price / 1M | — | — | — |
| Cached input / 1M | — | — | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | — | — | — |
| Modalities | text · vision | text | Amazon Nova 2 Lite |
| Released | 2025-12-02 | 2025-04-30 | — |
| Cited benchmark scores | 12 | 4 | — |
Reliability
Provider status
Last 60 days · refreshed daily on this page · live on /statusMore matchups
Amazon Nova 2 Lite vs …
Models sharing the most benchmarksMore matchups
Phi-4 Reasoning vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Amazon Nova 2 Lite and Phi-4 Reasoning on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.