Kimi K1.5 Instruct vs MiniMax-M2: which model should you choose?
We compared Kimi K1.5 Instruct and MiniMax-M2 on price, context window and 12 benchmarks. MiniMax-M2 has the stronger average benchmark score; MiniMax-M2 is cheaper on a blended 3:1 basis.
Kimi K1.5 Instruct
Kimi K1.5 Instruct is a multimodal model from Moonshot AI. It offers a 200,000-token context window, costs $5.15 per million input tokens and $28.20 per million output tokens, and averages 80.5 across the index's benchmark suite.
- Input
- $5.15 / 1M
- Output
- $28.20 / 1M
- Context
- 200,000
- Avg score
- 80.5
- Open weights
- Yes
MiniMax-M2
MiniMax-M2 is a multimodal model from MiniMax. It offers a 200,000-token context window, costs $5.19 per million input tokens and $22.73 per million output tokens, and averages 80.6 across the index's benchmark suite.
- Input
- $5.19 / 1M
- Output
- $22.73 / 1M
- Context
- 200,000
- Avg score
- 80.6
- Open weights
- Yes
Specification comparison
| Spec | Kimi K1.5 Instruct | MiniMax-M2 | Edge |
|---|---|---|---|
| Input price / 1M | $5.15 | $5.19 | A |
| Output price / 1M | $28.20 | $22.73 | B |
| Context window | 200,000 | 200,000 | Tie |
| Max output | 4,096 | 65,536 | B |
| Average score | 80.5 | 80.6 | B |
| Arena Elo | 1431.8 | 1423.8 | A |
| Open weights | Yes | Yes | Tie |
Benchmark comparison
Kimi K1.5 Instruct wins 8, MiniMax-M2 wins 4, and 0 benchmarks are tied. The table shows the modelled score for each model and the gap.
| Benchmark | Kimi K1.5 Instruct | MiniMax-M2 | Gap | Winner |
|---|---|---|---|---|
| MMLU-Pro | 83.0 | 81.8 | 1.2 | Kimi K1.5 Instruct |
| GPQA Diamond | 77.4 | 75.4 | 2.0 | Kimi K1.5 Instruct |
| SWE-bench Verified | 67.8 | 69.9 | 2.1 | MiniMax-M2 |
| HumanEval | 84.0 | 88.0 | 4.0 | MiniMax-M2 |
| MATH-500 | 82.6 | 82.3 | 0.3 | Kimi K1.5 Instruct |
| AIME 2025 | 80.4 | 78.1 | 2.3 | Kimi K1.5 Instruct |
| LiveCodeBench | 71.1 | 73.4 | 2.3 | MiniMax-M2 |
| IFEval | 86.2 | 86.1 | 0.1 | Kimi K1.5 Instruct |
| TruthfulQA | 84.3 | 83.5 | 0.8 | Kimi K1.5 Instruct |
| HellaSwag | 90.2 | 90.4 | 0.2 | MiniMax-M2 |
| MMMU | 78.6 | 78.0 | 0.6 | Kimi K1.5 Instruct |
| Arena Elo | 1431.8 | 1423.8 | 8.0 | Kimi K1.5 Instruct |
Price analysis
On a blended 3:1 input-to-output basis, Kimi K1.5 Instruct costs about $10.91 per million tokens and MiniMax-M2 costs about $9.57. MiniMax-M2 is roughly 12% cheaper than Kimi K1.5 Instruct. At high volume, that gap compounds quickly, so price should carry real weight unless the quality difference is decisive.
Both models list a 200,000-token context window. That is enough for long documents, large codebases and multi-step agent traces.
Verdict
Across 12 shared benchmarks, Kimi K1.5 Instruct wins 8, MiniMax-M2 wins 4, and 0 are tied. On the index average, MiniMax-M2 leads Kimi K1.5 Instruct by 0.1 points (80.6 versus 80.5).
The price picture favours MiniMax-M2. The context picture is effectively a tie. Those two dimensions often matter more than a few benchmark points once a model is good enough for the task.
If you need a single default, start with MiniMax-M2 for quality-sensitive work, and check whether its price holds at your expected volume before committing.
Choose Kimi K1.5 Instruct if
- You prefer Moonshot AI's tooling or already use it.
Choose MiniMax-M2 if
- You want the higher average benchmark score in this index.
- Unit cost matters and you want the cheaper blended price.
- You need the larger 65,536-token output limit.
FAQ
Is Kimi K1.5 Instruct better than MiniMax-M2?
On the index average, MiniMax-M2 leads Kimi K1.5 Instruct by 0.1 points (80.6 versus 80.5). The better choice depends on price, context and your task. Kimi K1.5 Instruct wins 8 benchmarks, MiniMax-M2 wins 4.
Which is cheaper, Kimi K1.5 Instruct or MiniMax-M2?
MiniMax-M2 is cheaper on a blended 3:1 basis at about $9.57 per million tokens, versus $10.91 for Kimi K1.5 Instruct.
Which has the larger context window?
Both models list a 200,000-token context window.
Are these models open weights?
Kimi K1.5 Instruct is open weights; MiniMax-M2 is open weights.
Which should I choose for production?
If you need a single default, start with MiniMax-M2 for quality-sensitive work, and check whether its price holds at your expected volume before committing.