MiniMax M2.1 vs Step-3.5-Flash: Specs & Benchmark Comparison
| Characteristic | MiniMax M2.1 | Step-3.5-Flash |
|---|---|---|
| Company | MiniMax | StepFun |
| Release Date | December 22, 2025 | February 1, 2026 |
| Parameters | — | 196B |
| Multimodal | No | Yes |
| Context (input) | 1.0M | 66K |
| Context (output) | 100K | 8K |
| Input Price / 1M | $0.30 | $0.10 |
| Output Price / 1M | $1.20 | $0.40 |
| Average Score | 0.9 | 0.8 |
Verdict
Step-3.5-Flash leads in 2 out of 4 comparison categories.
Overall Performance
Both models show comparable average scores: MiniMax M2.1 — 0.9, Step-3.5-Flash — 0.8.
API Cost
Step-3.5-Flash is 3.0x cheaper: input $0.10/1M vs $0.30/1M tokens.
Context Window
MiniMax M2.1 supports a larger context: 1M vs 66K tokens.
Recency
Step-3.5-Flash is newer: released 2/1/2026 vs 12/22/2025.
More About These Models
Related Comparisons
The MiniMax M2.1 and Step-3.5-Flash comparison is updated for 2026. Data includes benchmark results, API pricing, context window size and other specifications. For more detailed information, visit the MiniMax M2.1 or Step-3.5-Flash page. See also the complete list of AI model comparisons.