Llama 4 Maverick 17B 128E Instruct vs MiniMax M3 (Consensus Protocol)
MiniMax M3 (Consensus Protocol) is cheaper on output tokens, while MiniMax M3 (Consensus Protocol) offers a larger context window. Choose Llama 4 Maverick 17B 128E Instruct or MiniMax M3 (Consensus Protocol) based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | Llama 4 Maverick 17B 128E Instruct | MiniMax M3 (Consensus Protocol) |
|---|---|---|
| Provider | Vertex | LLM Gateway |
| Input / 1M tokens | $0.35 | $0.28 |
| Output / 1M tokens | $1.15 | $1.10 |
| Context window | 524K | 1M |
| Parameters | — | — |
| Open weights | Yes | Yes |
| Released | Apr 2025 | Jun 2026 |
| LMArena (Chatbot Arena) Elo score | 1287 | — |
| LMArena Vision Elo score | 1142 | — |
FAQ
Is Llama 4 Maverick 17B 128E Instruct or MiniMax M3 (Consensus Protocol) cheaper?
MiniMax M3 (Consensus Protocol) is cheaper on output tokens ($1.10 vs $1.15 per 1M).
Which has the larger context window, Llama 4 Maverick 17B 128E Instruct or MiniMax M3 (Consensus Protocol)?
MiniMax M3 (Consensus Protocol) has the larger context window (1M tokens).
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.