Model comparison
Mercury 2.5 vs MiniMax-M3
MiniMax-M3 is the stronger model overall, scoring 43.8 to 33.5 on the Noometry Index. Mercury 2.5 costs 7.8× less per token, which makes it the better buy when MiniMax-M3's lead doesn't matter for your workload.
Last verified . 4 shared benchmarks.
MiniMaxSummary
- They share 4 benchmarks with published results for both. Mercury 2.5 scores higher in 0 categories and MiniMax-M3 in 3 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in math, where MiniMax-M3 leads 40.0 to 23.3.
- The biggest single-benchmark swing is ProofBench: 3% for Mercury 2.5 and 18% for MiniMax-M3.
- Mercury 2.5 is cheaper at $0.04 / $0.15 per million input/output tokens, against $0.30 / $1.20 for MiniMax-M3.
- MiniMax-M3 accepts more context: 1M tokens versus 260K.
- MiniMax-M3 has downloadable open weights; the other is API-only.
Side by side
| Mercury 2.5 | MiniMax-M3 | |
|---|---|---|
| Provider | Inception | MiniMax |
| Noometry Index | 33.5 | 43.8 |
| Released | 2026-09-08 | 2026-06-01 |
| Weights | Proprietary | Open |
| Context window | 260K | 1M |
| Max output | 66K | 512K |
| Input $ / M tokens | $0.04 | $0.30 |
| Output $ / M tokens | $0.15 | $1.20 |
| Results tracked | 4 | 41 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M3 leads
Mercury 2.5: 39.5 (#156), MiniMax-M3: 41.8 (#118)
| Benchmark | Mercury 2.5 | MiniMax-M3 |
|---|---|---|
| SciCode | 38.5% | 47.1% |
| ALE-Bench | 301.65 | 640.02 |
| FrontierCode | — | 14.7% |
| LMArena WebDev | — | 1482 |
| LMArena Coding | — | 1469 |
Agentic & Tool Use Not comparable
Mercury 2.5: —, MiniMax-M3: 22.6 (#130)
| Benchmark | Mercury 2.5 | MiniMax-M3 |
|---|---|---|
| APEX-Agents | — | 37.7% |
| OSWorld 2.0 | — | 4.6% |
| GBAEval | — | 0.9% |
| Vending-Bench 2 | — | 2,158 |
Reasoning MiniMax-M3 leads
Mercury 2.5: 22.4 (#193), MiniMax-M3: 30.1 (#87)
| Benchmark | Mercury 2.5 | MiniMax-M3 |
|---|---|---|
| CritPt | 0% | 3.7% |
| SimpleBench | — | 45.8% |
| NYT Connections (extended) | — | 65.1% |
| Chess Puzzles | — | 14% |
| LMArena Hard Prompts | — | 1447 |
| Mystery Game Puzzles | — | 8% |
| DTBench | — | 78.9% |
| LMCA | — | 33.7% |
| Surface Evolver Bench | — | 55% |
| Epoch Capabilities Index | — | 146.95 |
| ForecastBench | — | 61.4 |
Math MiniMax-M3 leads
Mercury 2.5: 23.3 (#272), MiniMax-M3: 40.0 (#95)
| Benchmark | Mercury 2.5 | MiniMax-M3 |
|---|---|---|
| ProofBench | 3% | 18% |
| OTIS Mock AIME 2024-2025 | — | 71.1% |
| LMArena Math | — | 1429 |
Knowledge Not comparable
Mercury 2.5: —, MiniMax-M3: 58.4 (#35)
| Benchmark | Mercury 2.5 | MiniMax-M3 |
|---|---|---|
| GPQA Diamond | — | 90.9% |
| LMArena Expert | — | 1461 |
Multimodal Not comparable
Mercury 2.5: —, MiniMax-M3: 40.2 (#51)
| Benchmark | Mercury 2.5 | MiniMax-M3 |
|---|---|---|
| LMArena Vision | — | 1253 |
| LMArena Document | — | 1435 |
Multilingual Not comparable
Mercury 2.5: —, MiniMax-M3: 53.0 (#75)
| Benchmark | Mercury 2.5 | MiniMax-M3 |
|---|---|---|
| LMArena Non-English | — | 1420 |
| LMArena Chinese | — | 1463 |
| LMArena French | — | 1447 |
| LMArena German | — | 1426 |
| LMArena Japanese | — | 1381 |
| LMArena Korean | — | 1372 |
| LMArena Russian | — | 1428 |
| LMArena Spanish | — | 1432 |
Instruction Following Not comparable
Mercury 2.5: —, MiniMax-M3: 75.5 (#62)
| Benchmark | Mercury 2.5 | MiniMax-M3 |
|---|---|---|
| LMArena Instruction Following | — | 1433 |
Long Context Not comparable
Mercury 2.5: —, MiniMax-M3: 44.2 (#72)
| Benchmark | Mercury 2.5 | MiniMax-M3 |
|---|---|---|
| LMArena Longer Query | — | 1445 |
Writing & Preference Not comparable
Mercury 2.5: —, MiniMax-M3: 62.1 (#83)
| Benchmark | Mercury 2.5 | MiniMax-M3 |
|---|---|---|
| LMArena Text | — | 1433 |
| LMArena Creative Writing | — | 1404 |
| EQ-Bench 4 | — | 1150 |
| LMArena Multi-Turn | — | 1442 |
Frequently asked questions
Is Mercury 2.5 better than MiniMax-M3?
MiniMax-M3 is the stronger model overall, scoring 43.8 to 33.5 on the Noometry Index. Mercury 2.5 costs 7.8× less per token, which makes it the better buy when MiniMax-M3's lead doesn't matter for your workload.
Which is cheaper, Mercury 2.5 or MiniMax-M3?
Mercury 2.5 is cheaper. It lists at $0.04 per million input tokens and $0.15 per million output tokens; MiniMax-M3 lists at $0.30 and $1.20.
Is Mercury 2.5 or MiniMax-M3 better for coding?
MiniMax-M3 scores higher on coding benchmarks: 41.8 versus 39.5 in the Noometry coding category.
Which has the bigger context window?
MiniMax-M3 does, with 1M tokens against 260K.
How many benchmarks do Mercury 2.5 and MiniMax-M3 share?
4 benchmarks have published results for both models. Mercury 2.5 has 4 scored results on Noometry and MiniMax-M3 has 41.