$ diff qwen4-max mistral-large-4
Qwen4-Max vs Mistral Large 4
Values come from Qwen4-Max and Mistral Large 4, where each is cited to its source. This page states no benchmark result and ranks neither model — it puts two published specifications next to each other. Where a lab has not published a figure, the row says so rather than guessing.
Alibaba QwenQwen4-Max
- context
- 1M
- weights
- —
- $/M in
- $1.5
- $/M out
- $4.5
- context
- 1M
- weights
- open
- $/M in
- $0.7
- $/M out
- $2.1
What actually differs
- Context window
- Identical — both accept 1M tokens, so context length is not a reason to pick either.
- Input price
- Mistral Large 4 at $0.7/M against $1.5/M — 2.1× cheaper to feed. Standard rates: one of these labs also quotes a lower cached-input tier, which applies only when a prefix is reused — see the full spec below.
- Output price
- Mistral Large 4 at $2.1/M against $4.5/M — 2.1× cheaper to generate. Output dominates the bill on most agentic workloads, where the model writes far more than it reads.
- Recency
- Qwen4-Max shipped 44 days after Mistral Large 4 (2026-11-19 vs 2026-10-06).
Full spec
| Attribute | Qwen4-Max | Mistral Large 4 |
|---|---|---|
| Developer | Alibaba Qwen | Mistral AI |
| Released | 2026-11-19 | 2026-10-06 |
| Context window | 1,000,000 tokens | 1M |
| Pricing | Singapore International: $1.5/M input · $4.5/M output, thinking and non-thinking | $0.70/M input · $0.07/M cached input · $2.10/M output (2026-10-29 announcement) |
| License | not recorded | Mistral Ecosystem Open-Weight license; commercial Model-as-a-Service requires a separate license |
| Availability | Alibaba Cloud Model Studio API; model ID `qwen4-max` | Mistral Studio/API, Hugging Face weights for self-hosting, Microsoft Foundry, NVIDIA endpoints |