The LLM benchmark aggregator.

Amazon Nova 2 Lite

Amazon

Strongest in Factual grounding (#10 of 102), weakest in Agents (#215 of 289). Above par in 4 of 9 scopes. Among the models it meets almost everywhere, it finishes behind GPT-5.6 Terra and Claude Sonnet 5 and ahead of Gemini 2.5 Pro and DeepSeek R1.

§ 1 · Profile

What it is good at

Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.

Safety53.9−7.5#72/3311/3

Factual grounding62.4−8.4#10/1021/1

Knowledge38−24.7#184/1911/3

Document parsing37.2−26.4#70/771/1

Human preference50.5−18.1#193/3501/1

Agents44.8−21.1#215/2892/6

Tool use42.2−31.8#57/811/1

Knowledge work42.7−29.1#128/1981/2

§ 2 · Head to head

What it beats, and what beats it

The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for Amazon Nova 2 Lite, left for the other.

Llama 3.1 8B Instructwon 6 · lost 0

Ministral 8Bwon 6 · lost 0

Gemma 3 4B ITwon 6 · lost 0

NVIDIA-Nemotron-3-Nano-30B-A3B (BF16)won 6 · lost 0

GPT-4o miniwon 5 · lost 1

Llama 4 Scout Instructwon 6 · lost 2

Llama 4 Maverick Instructwon 6 · lost 2

GPT-4.1 nanowon 5 · lost 2

Gemma 4 E4Bwon 4 · lost 2

Mistral Mediumwon 4 · lost 2

GPT OSS 20Bwon 4 · lost 2

Mercury 2won 4 · lost 2

Trinity Largewon 4 · lost 2

Llama 3.3 70B Instructwon 5 · lost 3

Gemma 3 27B ITwon 4 · lost 3

DeepSeek V3won 4 · lost 3

Mistral Large 3won 4 · lost 3

DeepSeek R1won 4 · lost 3

Gemini 2.5 Prowon 4 · lost 3

Claude 3.5 Haikuwon 3 · lost 3

Qwen3 Next 80B A3B Thinkingwon 3 · lost 3

Gemini 3.1 Flash-Litewon 4 · lost 4

Gemma 4 26B A4Bwon 4 · lost 4

Gemma 4 31Bwon 4 · lost 5

Mistral Largewon 3 · lost 4

GPT-4owon 3 · lost 4

Gemma 3 12B ITwon 3 · lost 4

Qwen3 32Bwon 3 · lost 4

NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)won 3 · lost 4

Gemini 3.1 Prowon 3 · lost 5

Phi 4won 2 · lost 4

Granite 4.2 30Bwon 2 · lost 4

O3 Miniwon 2 · lost 4

Qwen3.5 35B A3Bwon 2 · lost 4

Qwen3.5 122B A10Bwon 2 · lost 4

Gemini 3 Flashwon 2 · lost 4

Claude Opus 4.6won 2 · lost 4

Gemini 4 Argonwon 2 · lost 4

Command A (03-2025)won 2 · lost 5

NVIDIA Nemotron 3 Super 120B A12Bwon 2 · lost 5

GPT-4.1 miniwon 2 · lost 5

O4 Miniwon 2 · lost 5

Gemini 2.5 Flashwon 2 · lost 5

Kimi K2won 2 · lost 5

Grok 4.1 Fastwon 2 · lost 5

GLM-4.7won 2 · lost 5

GPT-5.4 miniwon 2 · lost 5

Kimi K2.5won 2 · lost 5

Kimi K2.6won 2 · lost 5

DeepSeek-V4-Pro (High, 20260813)won 2 · lost 5

Claude Sonnet 4.6won 2 · lost 5

Claude Opus 4.5 (High, 32K, 20251101)won 2 · lost 5

Gemini 3 Prowon 2 · lost 5

DeepSeek V3.2won 2 · lost 6

GLM-4.6won 2 · lost 6

GPT-6 Solwon 2 · lost 6

GPT-6 Astrawon 2 · lost 6

GPT-5.6 Solwon 2 · lost 6

Grok 4won 1 · lost 5

DeepSeek V4 Flashwon 1 · lost 5

Grok 4.3won 1 · lost 5

GLM-5.1won 1 · lost 5

MiMo V2.5 Prowon 1 · lost 5

Qwen3.7 Maxwon 1 · lost 5

Qwen3.8 Maxwon 1 · lost 5

GPT-5 nanowon 1 · lost 6

GPT OSS 120Bwon 1 · lost 6

Claude Sonnet 4won 1 · lost 6

GPT-4.1won 1 · lost 6

MiniMax M2.7won 1 · lost 6

Qwen3 235B A22B Instructwon 1 · lost 6

GPT-5won 1 · lost 6

GPT-5.2won 1 · lost 6

Inklingwon 1 · lost 6

MiniMax M3won 1 · lost 6

GPT-6 Lunawon 1 · lost 6

GPT-5.1won 1 · lost 6

GLM-5.3 Flashwon 1 · lost 6

GLM-5.2won 1 · lost 6

Gemini 3.5 Flashwon 1 · lost 6

Gemini 3.6 Flashwon 1 · lost 6

Gemini 3.8 Flashwon 1 · lost 6

GPT-5.6 Lunawon 1 · lost 7

Claude Sonnet 4.5 (High, 32K, 20250929)won 1 · lost 7

GPT-5.4won 1 · lost 7

GPT-5.5won 1 · lost 7

Claude Opus 4.7won 1 · lost 7

GPT-5 miniwon 1 · lost 8

Claude Haiku 4.5won 1 · lost 8

O3won 0 · lost 6

Grok 4.6won 0 · lost 6

Gemini 3.5 Flash-Litewon 0 · lost 6

Grok 4.5won 0 · lost 6

DeepSeek V4.1 Flashwon 0 · lost 6

Claude Sonnet 5.5won 0 · lost 6

GPT-6.1 Solwon 0 · lost 6

Gemini 3.7 Flashwon 0 · lost 6

Muse Spark 1.1won 0 · lost 6

Claude Opus 5won 0 · lost 6

Claude Fable 5.1won 0 · lost 6

Claude Fable 5won 0 · lost 6

Gemini 2.5 Flash-Litewon 0 · lost 7

Qwen3.8 27Bwon 0 · lost 7

Claude Opus 4.8won 0 · lost 7

Claude Opus 5.5won 0 · lost 7

GPT-5.4 nanowon 0 · lost 8

Claude Sonnet 5won 0 · lost 8

GPT-5.6 Terrawon 0 · lost 8

§ 3 · Sources

Where the numbers come from

5 publications, 10 figures. Every one links to the page it was read from.

LMArena Text ↗1 measureread 2026-10-21

LMArena Text 1335

GDPval-AA ↗1 measureread 2026-10-21

GDPval-AA 366

BFCL ↗1 measureread 2026-10-21

BFCL v4 27.1%

ParseBench ↗6 measuresread 2026-10-21

ParseBench · Overall 40.43%ParseBench · Tables 68.38%ParseBench · Charts 6.7%ParseBench · Content faithfulness 77.04%ParseBench · Semantic formatting 45.26%ParseBench · Visual grounding 4.78%

Vectara Hallucination Leaderboard ↗1 measureread 2026-10-21

Vectara · Factual consistency 94.9%

Badge

[![PublicAI Index](https://publicai.io/model-index/badge?model=amazon-nova-2-lite)](https://publicai.io/model-index/m/amazon-nova-2-lite)