Model comparison
Llama 3.2 3B vs MiniMax-M2.5
MiniMax-M2.5 is the stronger model overall, scoring 38.3 to 28.9 on the Noometry Index. Llama 3.2 3B costs 4.4× less per token, which makes it the better buy when MiniMax-M2.5's lead doesn't matter for your workload.
Last verified . 14 shared benchmarks.
Meta
MiniMaxSummary
- They share 14 benchmarks with published results for both. Llama 3.2 3B scores higher in 2 categories and MiniMax-M2.5 in 7 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where MiniMax-M2.5 leads 53.9 to 24.7.
- Llama 3.2 3B is cheaper at $0.05 / $0.33 per million input/output tokens, against $0.30 / $1.20 for MiniMax-M2.5.
- MiniMax-M2.5 accepts more context: 205K tokens versus 131K.
Side by side
| Llama 3.2 3B | MiniMax-M2.5 | |
|---|---|---|
| Provider | Meta | MiniMax |
| Noometry Index | 28.9 | 38.3 |
| Released | 2024-09-24 | 2026-02-12 |
| Weights | Open | Open |
| Context window | 131K | 205K |
| Max output | 118K | 131K |
| Input $ / M tokens | $0.05 | $0.30 |
| Output $ / M tokens | $0.33 | $1.20 |
| Results tracked | 18 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2.5 leads
Llama 3.2 3B: 27.6 (#319), MiniMax-M2.5: 48.1 (#58)
| Benchmark | Llama 3.2 3B | MiniMax-M2.5 |
|---|---|---|
| LMArena Coding | 1098 | 1381 |
| SWE-bench Verified (bash only) | — | 75.8% |
| LMArena WebDev | — | 1387 |
| SWE-bench Multilingual | — | 68.3% |
| BigCodeBench Instruct | 23.4% | — |
| BigCodeBench Complete | 28.3% | — |
| ALE-Bench | — | 618.17 |
Agentic & Tool Use MiniMax-M2.5 leads
Llama 3.2 3B: 20.1 (#143), MiniMax-M2.5: 30.4 (#77)
| Benchmark | Llama 3.2 3B | MiniMax-M2.5 |
|---|---|---|
| Terminal-Bench | — | 42.7% |
| Berkeley Function Calling Leaderboard | 21.9% | — |
| BALROG | 10.1% | — |
| Vending-Bench 2 | — | -23.16 |
Reasoning Llama 3.2 3B leads
Llama 3.2 3B: 21.0 (#228), MiniMax-M2.5: 17.5 (#292)
| Benchmark | Llama 3.2 3B | MiniMax-M2.5 |
|---|---|---|
| LMArena Hard Prompts | 1095 | 1372 |
| ARC-AGI-2 | — | 4.9% |
| Kagi LLM Benchmark | — | 55.2% |
| NYT Connections (extended) | — | 16.8% |
| ARC-AGI-1 | — | 63.7% |
| Epoch Capabilities Index | — | 146.68 |
Math Llama 3.2 3B leads
Llama 3.2 3B: 32.4 (#214), MiniMax-M2.5: 26.9 (#253)
| Benchmark | Llama 3.2 3B | MiniMax-M2.5 |
|---|---|---|
| LMArena Math | 1126 | 1378 |
| ProofBench | — | 4% |
Knowledge MiniMax-M2.5 leads
Llama 3.2 3B: 29.7 (#235), MiniMax-M2.5: 39.2 (#135)
| Benchmark | Llama 3.2 3B | MiniMax-M2.5 |
|---|---|---|
| LMArena Expert | 1090 | 1379 |
| Vectara Hallucination Rate | — | 9.1% |
Multilingual MiniMax-M2.5 leads
Llama 3.2 3B: 26.2 (#281), MiniMax-M2.5: 47.1 (#152)
| Benchmark | Llama 3.2 3B | MiniMax-M2.5 |
|---|---|---|
| LMArena Non-English | 1019 | 1338 |
| LMArena Chinese | 1017 | 1393 |
| LMArena German | 1056 | 1362 |
| LMArena Russian | 949 | 1358 |
| LMArena French | — | 1362 |
| LMArena Japanese | — | 1171 |
| LMArena Korean | — | 1232 |
| LMArena Spanish | — | 1354 |
Instruction Following MiniMax-M2.5 leads
Llama 3.2 3B: 56.0 (#275), MiniMax-M2.5: 71.5 (#148)
| Benchmark | Llama 3.2 3B | MiniMax-M2.5 |
|---|---|---|
| LMArena Instruction Following | 1089 | 1353 |
Long Context MiniMax-M2.5 leads
Llama 3.2 3B: 33.4 (#261), MiniMax-M2.5: 37.5 (#216)
| Benchmark | Llama 3.2 3B | MiniMax-M2.5 |
|---|---|---|
| LMArena Longer Query | 1100 | 1366 |
| CL-bench | — | 11.4% |
| CL-bench Life | — | 6.3% |
Writing & Preference MiniMax-M2.5 leads
Llama 3.2 3B: 24.7 (#307), MiniMax-M2.5: 53.9 (#153)
| Benchmark | Llama 3.2 3B | MiniMax-M2.5 |
|---|---|---|
| LMArena Text | 1110 | 1359 |
| LMArena Creative Writing | 1094 | 1331 |
| EQ-Bench Creative Writing | 595 | 1361 |
| LMArena Multi-Turn | 1105 | 1364 |
Frequently asked questions
Is Llama 3.2 3B better than MiniMax-M2.5?
MiniMax-M2.5 is the stronger model overall, scoring 38.3 to 28.9 on the Noometry Index. Llama 3.2 3B costs 4.4× less per token, which makes it the better buy when MiniMax-M2.5's lead doesn't matter for your workload.
Which is cheaper, Llama 3.2 3B or MiniMax-M2.5?
Llama 3.2 3B is cheaper. It lists at $0.05 per million input tokens and $0.33 per million output tokens; MiniMax-M2.5 lists at $0.30 and $1.20.
Is Llama 3.2 3B or MiniMax-M2.5 better for coding?
MiniMax-M2.5 scores higher on coding benchmarks: 48.1 versus 27.6 in the Noometry coding category.
Which has the bigger context window?
MiniMax-M2.5 does, with 205K tokens against 131K.
How many benchmarks do Llama 3.2 3B and MiniMax-M2.5 share?
14 benchmarks have published results for both models. Llama 3.2 3B has 18 scored results on Noometry and MiniMax-M2.5 has 33.