Model comparison
Inkling vs MiniMax-M2.5
Inkling is the stronger model overall, scoring 44.1 to 38.3 on the Noometry Index. MiniMax-M2.5 costs 4.9× less per token, which makes it the better buy when Inkling's lead doesn't matter for your workload.
Last verified . 24 shared benchmarks.
MiniMaxSummary
- They share 24 benchmarks with published results for both. Inkling scores higher in 7 categories and MiniMax-M2.5 in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Inkling leads 40.4 to 17.5.
- The biggest single-benchmark swing is ARC-AGI-2: 36.5% for Inkling and 4.9% for MiniMax-M2.5.
- MiniMax-M2.5 is cheaper at $0.30 / $1.20 per million input/output tokens, against $1.87 / $4.68 for Inkling.
- MiniMax-M2.5 accepts more context: 205K tokens versus 66K.
Side by side
| Inkling | MiniMax-M2.5 | |
|---|---|---|
| Provider | Thinking Machines Lab | MiniMax |
| Noometry Index | 44.1 | 38.3 |
| Released | 2026-07-15 | 2026-02-12 |
| Weights | Open | Open |
| Context window | 66K | 205K |
| Max output | 66K | 131K |
| Input $ / M tokens | $1.87 | $0.30 |
| Output $ / M tokens | $4.68 | $1.20 |
| Results tracked | 41 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2.5 leads
Inkling: 34.5 (#234), MiniMax-M2.5: 48.1 (#58)
| Benchmark | Inkling | MiniMax-M2.5 |
|---|---|---|
| LMArena WebDev | 1413 | 1387 |
| LMArena Coding | 1464 | 1381 |
| ALE-Bench | 946 | 618.17 |
| FrontierCode | 14% | — |
| SWE-bench Verified (bash only) | — | 75.8% |
| SWE-bench Multilingual | — | 68.3% |
| FrontierSWE | 4.1% | — |
| SciCode | 47% | — |
| WeirdML | 32.3% | — |
Agentic & Tool Use Too close to call
Inkling: 29.6 (#85), MiniMax-M2.5: 30.4 (#77)
| Benchmark | Inkling | MiniMax-M2.5 |
|---|---|---|
| Terminal-Bench | — | 42.7% |
| APEX-Agents | 33.8% | — |
| τ²-bench Banking | 25% | — |
| Vending-Bench 2 | — | -23.16 |
Reasoning Inkling leads
Inkling: 40.4 (#56), MiniMax-M2.5: 17.5 (#292)
| Benchmark | Inkling | MiniMax-M2.5 |
|---|---|---|
| ARC-AGI-2 | 36.5% | 4.9% |
| ARC-AGI-1 | 79.5% | 63.7% |
| LMArena Hard Prompts | 1451 | 1372 |
| Epoch Capabilities Index | 148.54 | 146.68 |
| SimpleBench | 50% | — |
| Kagi LLM Benchmark | — | 55.2% |
| NYT Connections (extended) | — | 16.8% |
| CritPt | 5.4% | — |
| Chess Puzzles | 21% | — |
| DTBench | 87.5% | — |
| LMCA | 37.6% | — |
Math Inkling leads
Inkling: 31.3 (#225), MiniMax-M2.5: 26.9 (#253)
| Benchmark | Inkling | MiniMax-M2.5 |
|---|---|---|
| ProofBench | 0% | 4% |
| LMArena Math | 1479 | 1378 |
| FrontierMath (Tiers 1-3) | 33.3% | — |
| FrontierMath Tier 4 | 4.9% | — |
| OTIS Mock AIME 2024-2025 | 88.9% | — |
Knowledge Inkling leads
Inkling: 55.1 (#49), MiniMax-M2.5: 39.2 (#135)
| Benchmark | Inkling | MiniMax-M2.5 |
|---|---|---|
| LMArena Expert | 1465 | 1379 |
| GPQA Diamond | 88.3% | — |
| SimpleQA Verified | 40.3% | — |
| Vectara Hallucination Rate | — | 9.1% |
Multilingual Inkling leads
Inkling: 54.0 (#52), MiniMax-M2.5: 47.1 (#152)
| Benchmark | Inkling | MiniMax-M2.5 |
|---|---|---|
| LMArena Non-English | 1434 | 1338 |
| LMArena Chinese | 1490 | 1393 |
| LMArena French | 1458 | 1362 |
| LMArena German | 1446 | 1362 |
| LMArena Japanese | 1429 | 1171 |
| LMArena Korean | 1404 | 1232 |
| LMArena Russian | 1429 | 1358 |
| LMArena Spanish | 1448 | 1354 |
Instruction Following Inkling leads
Inkling: 75.1 (#71), MiniMax-M2.5: 71.5 (#148)
| Benchmark | Inkling | MiniMax-M2.5 |
|---|---|---|
| LMArena Instruction Following | 1426 | 1353 |
Long Context Inkling leads
Inkling: 43.8 (#86), MiniMax-M2.5: 37.5 (#216)
| Benchmark | Inkling | MiniMax-M2.5 |
|---|---|---|
| LMArena Longer Query | 1434 | 1366 |
| CL-bench | — | 11.4% |
| CL-bench Life | — | 6.3% |
Writing & Preference Inkling leads
Inkling: 65.2 (#51), MiniMax-M2.5: 53.9 (#153)
| Benchmark | Inkling | MiniMax-M2.5 |
|---|---|---|
| LMArena Text | 1441 | 1359 |
| LMArena Creative Writing | 1387 | 1331 |
| EQ-Bench Creative Writing | 1611 | 1361 |
| LMArena Multi-Turn | 1436 | 1364 |
| EQ-Bench 4 | 1226 | — |
Frequently asked questions
Is Inkling better than MiniMax-M2.5?
Inkling is the stronger model overall, scoring 44.1 to 38.3 on the Noometry Index. MiniMax-M2.5 costs 4.9× less per token, which makes it the better buy when Inkling's lead doesn't matter for your workload.
Which is cheaper, Inkling or MiniMax-M2.5?
MiniMax-M2.5 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; Inkling lists at $1.87 and $4.68.
Is Inkling or MiniMax-M2.5 better for coding?
MiniMax-M2.5 scores higher on coding benchmarks: 48.1 versus 34.5 in the Noometry coding category.
Which has the bigger context window?
MiniMax-M2.5 does, with 205K tokens against 66K.
How many benchmarks do Inkling and MiniMax-M2.5 share?
24 benchmarks have published results for both models. Inkling has 41 scored results on Noometry and MiniMax-M2.5 has 33.