# GPT-5.1 vs Nova 2 Lite

> GPT-5.1 is the stronger model overall, scoring 49.0 to 39.7 on the Noometry Index. Nova 2 Lite costs 4.0× less per token, which makes it the better buy when GPT-5.1's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gpt-5-1-vs-nova-2-lite
- Last updated: 2026-10-11
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. GPT-5.1 scores higher in 9 categories and Nova 2 Lite in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.1 leads 52.2 to 37.5.
- The biggest single-benchmark swing is Vectara Hallucination Rate: 10.9% for GPT-5.1 and 5.1% for Nova 2 Lite.
- Nova 2 Lite is cheaper at $0.30 / $2.50 per million input/output tokens, against $1.25 / $10 for GPT-5.1.
- Nova 2 Lite accepts more context: 1M tokens versus 400K.

## Snapshot

| | GPT-5.1 | Nova 2 Lite |
|---|---|---|
| Provider | OpenAI | Amazon |
| Noometry Index | 49.0 | 39.7 |
| Rank | 53 | 161 |
| Context | 400K | 1M |
| Input $/M | $1.25 | $0.30 |
| Output $/M | $10 | $2.50 |
| Weights | Proprietary | Proprietary |

## Coding

- GPT-5.1: 46.4 (#66)
- Nova 2 Lite: 40.7 (#134)

| Benchmark | GPT-5.1 | Nova 2 Lite |
|---|---|---|
| LMArena Coding | 1454 | 1385 |
| SWE-bench Verified | 68% | — |
| SWE-bench Verified (bash only) | 66% | — |
| LMArena WebDev | 1395 | — |
| SciCode | 43.3% | — |
| GSO | 13.7% | — |
| WeirdML | 60.8% | — |
| LiveBench Coding | 72.5% | — |
| ALE-Bench | 1,192 | — |

## Agentic & Tool Use

- GPT-5.1: 32.7 (#60)
- Nova 2 Lite: 24.1 (#121)

| Benchmark | GPT-5.1 | Nova 2 Lite |
|---|---|---|
| Terminal-Bench | 47.6% | — |
| Berkeley Function Calling Leaderboard | — | 27.1% |
| DeepResearch Bench | 42.8% | — |
| LMArena Search | 1199 | — |
| Vending-Bench 2 | 1,473 | — |

## Reasoning

- GPT-5.1: 39.8 (#58)
- Nova 2 Lite: 27.5 (#118)

| Benchmark | GPT-5.1 | Nova 2 Lite |
|---|---|---|
| LMArena Hard Prompts | 1457 | 1364 |
| ARC-AGI-2 | 17.6% | — |
| SimpleBench | 53.2% | — |
| ARC-AGI-1 | 72.8% | — |
| CritPt | 4.9% | — |
| Chess Puzzles | 32% | — |
| EnigmaEval | 11.2% | — |
| LiveBench Reasoning | 95.8% | — |
| Mystery Game Puzzles | 19% | — |
| DTBench | 90.1% | — |
| LiveBench Data Analysis | 72.1% | — |
| LMCA | 43.9% | — |
| Epoch Capabilities Index | 149.64 | — |
| ForecastBench | 58.1 | — |
| LiveBench | 78.8% | — |

## Math

- GPT-5.1: 52.2 (#51)
- Nova 2 Lite: 37.5 (#156)

| Benchmark | GPT-5.1 | Nova 2 Lite |
|---|---|---|
| LMArena Math | 1447 | 1359 |
| OTIS Mock AIME 2024-2025 | 88.6% | — |
| Omni-MATH | 46.4% | — |
| LiveBench Math | 94.5% | — |
| FrontierMath (Feb 2025 set) | 31% | — |
| FrontierMath Tier 4 (v1) | 12.5% | — |

## Knowledge

- GPT-5.1: 50.6 (#71)
- Nova 2 Lite: 43.0 (#94)

| Benchmark | GPT-5.1 | Nova 2 Lite |
|---|---|---|
| Vectara Hallucination Rate | 10.9% | 5.1% |
| LMArena Expert | 1470 | 1358 |
| GPQA Diamond | 87.6% | — |
| Humanity's Last Exam | 23.7% | — |
| SimpleQA Verified | 48% | — |
| MMLU-Pro | 57.9% | — |
| GPQA (HELM) | 44.2% | — |

## Multimodal

- GPT-5.1: 44.8 (#19)
- Nova 2 Lite: —

| Benchmark | GPT-5.1 | Nova 2 Lite |
|---|---|---|
| LMArena Vision | 1250 | — |
| VPCT | 58.7% | — |
| LMArena Document | 1403 | — |

## Multilingual

- GPT-5.1: 53.8 (#56)
- Nova 2 Lite: 47.1 (#153)

| Benchmark | GPT-5.1 | Nova 2 Lite |
|---|---|---|
| LMArena Non-English | 1431 | 1337 |
| LMArena Chinese | 1495 | 1364 |
| LMArena French | 1450 | 1381 |
| LMArena German | 1438 | 1343 |
| LMArena Japanese | 1453 | 1271 |
| LMArena Korean | 1401 | 1284 |
| LMArena Russian | 1435 | 1343 |
| LMArena Spanish | 1433 | 1373 |

## Instruction Following

- GPT-5.1: 83.9 (#1)
- Nova 2 Lite: 70.5 (#161)

| Benchmark | GPT-5.1 | Nova 2 Lite |
|---|---|---|
| LMArena Instruction Following | 1443 | 1335 |
| LiveBench Instruction Following | 93.3% | — |
| IFEval | 93.5% | — |

## Long Context

- GPT-5.1: 47.6 (#14)
- Nova 2 Lite: 40.6 (#150)

| Benchmark | GPT-5.1 | Nova 2 Lite |
|---|---|---|
| LMArena Longer Query | 1447 | 1335 |
| CL-bench | 23.7% | — |
| CL-bench Life | 17.3% | — |

## Writing & Preference

- GPT-5.1: 64.5 (#55)
- Nova 2 Lite: 53.9 (#154)

| Benchmark | GPT-5.1 | Nova 2 Lite |
|---|---|---|
| LMArena Text | 1443 | 1362 |
| LMArena Creative Writing | 1427 | 1291 |
| LMArena Multi-Turn | 1450 | 1338 |
| WildBench | 86.3% | — |
| LiveBench Language | 80.2% | — |

## FAQ

### Is GPT-5.1 better than Nova 2 Lite?

GPT-5.1 is the stronger model overall, scoring 49.0 to 39.7 on the Noometry Index. Nova 2 Lite costs 4.0× less per token, which makes it the better buy when GPT-5.1's lead doesn't matter for your workload.

### Which is cheaper, GPT-5.1 or Nova 2 Lite?

Nova 2 Lite is cheaper. It lists at $0.30 per million input tokens and $2.50 per million output tokens; GPT-5.1 lists at $1.25 and $10.

### Is GPT-5.1 or Nova 2 Lite better for coding?

GPT-5.1 scores higher on coding benchmarks: 46.4 versus 40.7 in the Noometry coding category.

### Which has the bigger context window?

Nova 2 Lite does, with 1M tokens against 400K.

### How many benchmarks do GPT-5.1 and Nova 2 Lite share?

18 benchmarks have published results for both models. GPT-5.1 has 63 scored results on Noometry and Nova 2 Lite has 19.