# ERNIE 5.1 vs MiniMax-M2.7

> ERNIE 5.1 is the stronger model overall, scoring 43.8 to 37.7 on the Noometry Index.

- Canonical page: https://noometry.com/compare/ernie-5-1-vs-minimax-m2-7
- Last updated: 2026-10-10
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. ERNIE 5.1 scores higher in 8 categories and MiniMax-M2.7 in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where ERNIE 5.1 leads 40.3 to 25.9.
- MiniMax-M2.7 has downloadable open weights; the other is API-only.

## Snapshot

| | ERNIE 5.1 | MiniMax-M2.7 |
|---|---|---|
| Provider | Baidu | MiniMax |
| Noometry Index | 43.8 | 37.7 |
| Rank | 83 | 196 |
| Context | — | 205K |
| Input $/M | — | $0.30 |
| Output $/M | — | $1.20 |
| Weights | Proprietary | Open |

## Coding

- ERNIE 5.1: 44.0 (#76)
- MiniMax-M2.7: 41.8 (#120)

| Benchmark | ERNIE 5.1 | MiniMax-M2.7 |
|---|---|---|
| LMArena Coding | 1488 | 1454 |
| LMArena WebDev | — | 1398 |
| SciCode | — | 47% |
| WeirdML | — | 37% |
| ALE-Bench | — | 599.25 |

## Agentic & Tool Use

- ERNIE 5.1: —
- MiniMax-M2.7: 25.1 (#111)

| Benchmark | ERNIE 5.1 | MiniMax-M2.7 |
|---|---|---|
| Terminal-Bench | — | 45.1% |
| ExploitBench | — | 13.3% |
| GBAEval | — | 0% |
| LMArena Search | 1227 | — |

## Reasoning

- ERNIE 5.1: 21.9 (#211)
- MiniMax-M2.7: 19.7 (#253)

| Benchmark | ERNIE 5.1 | MiniMax-M2.7 |
|---|---|---|
| NYT Connections (extended) | 23.4% | 24.7% |
| LMArena Hard Prompts | 1481 | 1422 |
| CritPt | — | 0.6% |
| Thematic Generalization | — | 39.3% |
| Epoch Capabilities Index | — | 145.85 |

## Math

- ERNIE 5.1: 40.3 (#92)
- MiniMax-M2.7: 25.9 (#263)

| Benchmark | ERNIE 5.1 | MiniMax-M2.7 |
|---|---|---|
| LMArena Math | 1481 | 1420 |
| ProofBench | — | 3% |

## Knowledge

- ERNIE 5.1: 41.9 (#102)
- MiniMax-M2.7: 37.7 (#152)

| Benchmark | ERNIE 5.1 | MiniMax-M2.7 |
|---|---|---|
| LMArena Expert | 1492 | 1444 |
| Vectara Hallucination Rate | — | 12.9% |

## Multilingual

- ERNIE 5.1: 55.5 (#29)
- MiniMax-M2.7: 50.3 (#123)

| Benchmark | ERNIE 5.1 | MiniMax-M2.7 |
|---|---|---|
| LMArena Non-English | 1454 | 1382 |
| LMArena Chinese | 1508 | 1441 |
| LMArena French | 1488 | 1421 |
| LMArena German | 1470 | 1398 |
| LMArena Japanese | 1422 | 1262 |
| LMArena Korean | 1427 | 1313 |
| LMArena Russian | 1459 | 1383 |
| LMArena Spanish | 1473 | 1403 |

## Instruction Following

- ERNIE 5.1: 76.7 (#37)
- MiniMax-M2.7: 74.1 (#103)

| Benchmark | ERNIE 5.1 | MiniMax-M2.7 |
|---|---|---|
| LMArena Instruction Following | 1460 | 1405 |

## Long Context

- ERNIE 5.1: 44.7 (#59)
- MiniMax-M2.7: 43.3 (#99)

| Benchmark | ERNIE 5.1 | MiniMax-M2.7 |
|---|---|---|
| LMArena Longer Query | 1462 | 1419 |

## Writing & Preference

- ERNIE 5.1: 65.1 (#52)
- MiniMax-M2.7: 58.9 (#112)

| Benchmark | ERNIE 5.1 | MiniMax-M2.7 |
|---|---|---|
| LMArena Text | 1468 | 1405 |
| LMArena Creative Writing | 1441 | 1354 |
| LMArena Multi-Turn | 1471 | 1412 |

## FAQ

### Is ERNIE 5.1 better than MiniMax-M2.7?

ERNIE 5.1 is the stronger model overall, scoring 43.8 to 37.7 on the Noometry Index.

### Is ERNIE 5.1 or MiniMax-M2.7 better for coding?

ERNIE 5.1 scores higher on coding benchmarks: 44.0 versus 41.8 in the Noometry coding category.

### How many benchmarks do ERNIE 5.1 and MiniMax-M2.7 share?

18 benchmarks have published results for both models. ERNIE 5.1 has 19 scored results on Noometry and MiniMax-M2.7 has 30.