# Gemini 2.5 Flash-Lite vs Muse Spark

> Muse Spark is the stronger model overall, scoring 50.6 to 37.0 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gemini-2-5-flash-lite-vs-muse-spark
- Last updated: 2026-10-10
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. Gemini 2.5 Flash-Lite scores higher in 0 categories and Muse Spark in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Muse Spark leads 65.7 to 32.5.

## Snapshot

| | Gemini 2.5 Flash-Lite | Muse Spark |
|---|---|---|
| Provider | Google | Meta |
| Noometry Index | 37.0 | 50.6 |
| Rank | 211 | 46 |
| Context | 1.05M | — |
| Input $/M | $0.10 | — |
| Output $/M | $0.40 | — |
| Weights | Proprietary | Proprietary |

## Coding

- Gemini 2.5 Flash-Lite: 38.5 (#173)
- Muse Spark: 46.2 (#69)

| Benchmark | Gemini 2.5 Flash-Lite | Muse Spark |
|---|---|---|
| LMArena Coding | 1373 | 1481 |
| SciCode | — | 51.5% |
| WeirdML | 35.2% | — |
| ALE-Bench | 325.9 | — |

## Agentic & Tool Use

- Gemini 2.5 Flash-Lite: 28.0 (#96)
- Muse Spark: —

| Benchmark | Gemini 2.5 Flash-Lite | Muse Spark |
|---|---|---|
| Berkeley Function Calling Leaderboard | 36.9% | — |

## Reasoning

- Gemini 2.5 Flash-Lite: 22.2 (#205)
- Muse Spark: 35.9 (#67)

| Benchmark | Gemini 2.5 Flash-Lite | Muse Spark |
|---|---|---|
| LMArena Hard Prompts | 1377 | 1474 |
| Epoch Capabilities Index | 133.94 | 152.04 |
| Kagi LLM Benchmark | 40.5% | — |
| CritPt | — | 11.3% |
| DTBench | 62.8% | — |
| LMCA | 18.1% | — |

## Math

- Gemini 2.5 Flash-Lite: 38.0 (#144)
- Muse Spark: 47.8 (#66)

| Benchmark | Gemini 2.5 Flash-Lite | Muse Spark |
|---|---|---|
| LMArena Math | 1373 | 1455 |
| OTIS Mock AIME 2024-2025 | — | 88.9% |
| ProofBench | — | 17% |
| Omni-MATH | 48% | — |
| FrontierMath (Feb 2025 set) | — | 39% |
| FrontierMath Tier 4 (v1) | — | 14.6% |

## Knowledge

- Gemini 2.5 Flash-Lite: 32.5 (#210)
- Muse Spark: 65.7 (#13)

| Benchmark | Gemini 2.5 Flash-Lite | Muse Spark |
|---|---|---|
| LMArena Expert | 1373 | 1457 |
| GPQA Diamond | — | 89.8% |
| Humanity's Last Exam | — | 40.6% |
| MMLU-Pro | 53.7% | — |
| Vectara Hallucination Rate | 3.3% | — |
| GPQA (HELM) | 30.9% | — |

## Multimodal

- Gemini 2.5 Flash-Lite: 29.1 (#114)
- Muse Spark: 43.4 (#24)

| Benchmark | Gemini 2.5 Flash-Lite | Muse Spark |
|---|---|---|
| LMArena Vision | 1198 | 1306 |
| VPCT | 30% | — |
| LMArena Document | — | 1444 |

## Multilingual

- Gemini 2.5 Flash-Lite: 49.3 (#134)
- Muse Spark: 56.1 (#24)

| Benchmark | Gemini 2.5 Flash-Lite | Muse Spark |
|---|---|---|
| LMArena Non-English | 1369 | 1464 |
| LMArena Chinese | 1404 | 1509 |
| LMArena French | 1388 | 1497 |
| LMArena German | 1389 | 1497 |
| LMArena Korean | 1360 | 1459 |
| LMArena Russian | 1373 | 1466 |
| LMArena Spanish | 1396 | 1472 |
| LMArena Japanese | 1359 | — |

## Instruction Following

- Gemini 2.5 Flash-Lite: 70.0 (#168)
- Muse Spark: 75.9 (#51)

| Benchmark | Gemini 2.5 Flash-Lite | Muse Spark |
|---|---|---|
| LMArena Instruction Following | 1367 | 1442 |
| IFEval | 81% | — |

## Long Context

- Gemini 2.5 Flash-Lite: 33.3 (#262)
- Muse Spark: 44.4 (#69)

| Benchmark | Gemini 2.5 Flash-Lite | Muse Spark |
|---|---|---|
| LMArena Longer Query | 1373 | 1451 |
| Fiction.LiveBench | 47.2% | — |

## Writing & Preference

- Gemini 2.5 Flash-Lite: 56.8 (#135)
- Muse Spark: 66.0 (#39)

| Benchmark | Gemini 2.5 Flash-Lite | Muse Spark |
|---|---|---|
| LMArena Text | 1379 | 1474 |
| LMArena Creative Writing | 1367 | 1459 |
| LMArena Multi-Turn | 1366 | 1477 |
| WildBench | 81.8% | — |

## FAQ

### Is Gemini 2.5 Flash-Lite better than Muse Spark?

Muse Spark is the stronger model overall, scoring 50.6 to 37.0 on the Noometry Index.

### Is Gemini 2.5 Flash-Lite or Muse Spark better for coding?

Muse Spark scores higher on coding benchmarks: 46.2 versus 38.5 in the Noometry coding category.

### How many benchmarks do Gemini 2.5 Flash-Lite and Muse Spark share?

18 benchmarks have published results for both models. Gemini 2.5 Flash-Lite has 33 scored results on Noometry and Muse Spark has 27.