# Amazon Nova Experimental Chat 10 20 vs Gemini 2.5 Flash

> Amazon Nova Experimental Chat 10 20 is the stronger model overall, scoring 41.5 to 39.0 on the Noometry Index.

- Canonical page: https://noometry.com/compare/amazon-nova-experimental-chat-10-20-vs-gemini-2-5-flash
- Last updated: 2026-10-10
- Shared benchmarks: 17

## Summary

- They share 17 benchmarks with published results for both. Amazon Nova Experimental Chat 10 20 scores higher in 4 categories and Gemini 2.5 Flash in 4 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Amazon Nova Experimental Chat 10 20 leads 27.3 to 18.2.

## Snapshot

| | Amazon Nova Experimental Chat 10 20 | Gemini 2.5 Flash |
|---|---|---|
| Provider | Amazon | Google |
| Noometry Index | 41.5 | 39.0 |
| Rank | 122 | 162 |
| Context | — | 1.05M |
| Input $/M | — | $0.30 |
| Output $/M | — | $2.50 |
| Weights | Proprietary | Proprietary |

## Coding

- Amazon Nova Experimental Chat 10 20: 41.0 (#123)
- Gemini 2.5 Flash: 35.5 (#210)

| Benchmark | Amazon Nova Experimental Chat 10 20 | Gemini 2.5 Flash |
|---|---|---|
| LMArena Coding | 1413 | 1424 |
| SWE-bench Verified (bash only) | — | 28.7% |
| Aider Polyglot | — | 55.1% |
| WeirdML | — | 41.9% |
| ALE-Bench | — | 661.88 |

## Agentic & Tool Use

- Amazon Nova Experimental Chat 10 20: —
- Gemini 2.5 Flash: 30.6 (#67)

| Benchmark | Amazon Nova Experimental Chat 10 20 | Gemini 2.5 Flash |
|---|---|---|
| Terminal-Bench | — | 17.1% |
| Berkeley Function Calling Leaderboard | — | 56.2% |
| TheAgentCompany | — | 41.1% |
| BALROG | — | 33.5% |
| Vending-Bench 2 | — | 548.84 |

## Reasoning

- Amazon Nova Experimental Chat 10 20: 27.3 (#110)
- Gemini 2.5 Flash: 18.2 (#271)

| Benchmark | Amazon Nova Experimental Chat 10 20 | Gemini 2.5 Flash |
|---|---|---|
| LMArena Hard Prompts | 1396 | 1422 |
| ARC-AGI-2 | — | 2.5% |
| SimpleBench | — | 41.2% |
| Kagi LLM Benchmark | — | 56.8% |
| ARC-AGI-1 | — | 33.3% |
| CritPt | — | 1.1% |
| EnigmaEval | — | 2.7% |
| DTBench | — | 76.5% |
| LMCA | — | 27.5% |
| Epoch Capabilities Index | — | 143.04 |
| ForecastBench | — | 60.6 |

## Math

- Amazon Nova Experimental Chat 10 20: 36.2 (#121)
- Gemini 2.5 Flash: 38.1 (#87)

| Benchmark | Amazon Nova Experimental Chat 10 20 | Gemini 2.5 Flash |
|---|---|---|
| LMArena Math | 1425 | 1415 |
| OTIS Mock AIME 2024-2025 | — | 73.1% |
| Omni-MATH | — | 38.5% |
| FrontierMath (Feb 2025 set) | — | 4.8% |
| FrontierMath Tier 4 (v1) | — | 4.2% |

## Knowledge

- Amazon Nova Experimental Chat 10 20: 37.8 (#142)
- Gemini 2.5 Flash: 36.1 (#162)

| Benchmark | Amazon Nova Experimental Chat 10 20 | Gemini 2.5 Flash |
|---|---|---|
| LMArena Expert | 1385 | 1426 |
| Humanity's Last Exam | — | 12.1% |
| MMLU-Pro | — | 63.9% |
| Confabulations | — | 16.8% |
| Vectara Hallucination Rate | — | 7.8% |
| GPQA (HELM) | — | 39% |

## Multimodal

- Amazon Nova Experimental Chat 10 20: —
- Gemini 2.5 Flash: 41.2 (#31)

| Benchmark | Amazon Nova Experimental Chat 10 20 | Gemini 2.5 Flash |
|---|---|---|
| LMArena Vision | — | 1254 |
| GeoBench | — | 76% |
| VPCT | — | 46.2% |
| SpatialViz-Bench | — | 36.9% |

## Multilingual

- Amazon Nova Experimental Chat 10 20: 49.7 (#127)
- Gemini 2.5 Flash: 52.5 (#83)

| Benchmark | Amazon Nova Experimental Chat 10 20 | Gemini 2.5 Flash |
|---|---|---|
| LMArena Non-English | 1372 | 1410 |
| LMArena Chinese | 1405 | 1450 |
| LMArena French | 1426 | 1433 |
| LMArena German | 1391 | 1419 |
| LMArena Japanese | 1355 | 1403 |
| LMArena Korean | 1327 | 1385 |
| LMArena Russian | 1371 | 1415 |
| LMArena Spanish | 1381 | 1421 |

## Instruction Following

- Amazon Nova Experimental Chat 10 20: 72.3 (#136)
- Gemini 2.5 Flash: 75.9 (#50)

| Benchmark | Amazon Nova Experimental Chat 10 20 | Gemini 2.5 Flash |
|---|---|---|
| LMArena Instruction Following | 1365 | 1405 |
| IFEval | — | 89.8% |

## Long Context

- Amazon Nova Experimental Chat 10 20: 41.7 (#128)
- Gemini 2.5 Flash: 47.5 (#16)

| Benchmark | Amazon Nova Experimental Chat 10 20 | Gemini 2.5 Flash |
|---|---|---|
| LMArena Longer Query | 1370 | 1419 |
| Fiction.LiveBench | — | 77.8% |

## Writing & Preference

- Amazon Nova Experimental Chat 10 20: 56.9 (#131)
- Gemini 2.5 Flash: 54.0 (#153)

| Benchmark | Amazon Nova Experimental Chat 10 20 | Gemini 2.5 Flash |
|---|---|---|
| LMArena Text | 1394 | 1417 |
| LMArena Creative Writing | 1318 | 1400 |
| LMArena Multi-Turn | 1364 | 1408 |
| Short-Story Creative Writing | — | 76.5% |
| EQ-Bench Creative Writing | — | 1137 |
| WildBench | — | 81.7% |

## FAQ

### Is Amazon Nova Experimental Chat 10 20 better than Gemini 2.5 Flash?

Amazon Nova Experimental Chat 10 20 is the stronger model overall, scoring 41.5 to 39.0 on the Noometry Index.

### Is Amazon Nova Experimental Chat 10 20 or Gemini 2.5 Flash better for coding?

Amazon Nova Experimental Chat 10 20 scores higher on coding benchmarks: 41.0 versus 35.5 in the Noometry coding category.

### How many benchmarks do Amazon Nova Experimental Chat 10 20 and Gemini 2.5 Flash share?

17 benchmarks have published results for both models. Amazon Nova Experimental Chat 10 20 has 17 scored results on Noometry and Gemini 2.5 Flash has 54.