Model comparison - Similar price, same tier
MiniMax-M3 vs Qwen3.8-Flash
MiniMax-M3 is from MiniMax and Qwen3.8-Flash is from Alibaba Cloud (Qwen). On list price, MiniMax-M3 charges 2x Qwen3.8-Flash's input price ($0.30 vs $0.15 per 1M tokens), and MiniMax-M3 charges 2.6x Qwen3.8-Flash's output price ($1.20 vs $0.47 per 1M tokens). MiniMax-M3 and Qwen3.8-Flash both have a 1,000,000-token context window; max output is 524,288 and 131,072 tokens per response. MiniMax-M3 was released on June 1, 2026 and Qwen3.8-Flash on August 26, 2026. MiniMax-M3 publishes open weights (MiniMax Community License (Hugging Face license tag: other / minimax-community)); Qwen3.8-Flash is closed-weights.
Last verified from each provider's pricing page. All sources
Side-by-side specs
| Spec | MiniMax-M3 | Qwen3.8-Flash |
|---|---|---|
| Provider | MiniMax | Alibaba Cloud (Qwen) |
| Input / 1M tokens | $0.30 | $0.15 |
| Output / 1M tokens | $1.20 | $0.47 |
| Cache read / 1M | $0.06 | $0.016 |
| Cache write / 1M | - | $0.20 |
| Long-context rates above | 512,000 tokens | - |
| Context window | 1,000,000 | 1,000,000 |
| Max output | 524,288 | 131,072 |
| Release date | ||
| Weights | Open | Closed |
| License | MiniMax Community License (Hugging Face license tag: other / minimax-community) | - |
| API model id | MiniMax-M3 | qwen3.8-flash |
| AI Gateway id | minimax/minimax-m3 | alibaba/qwen3.8-flash |
| Last verified |
Prices in USD per 1M tokens at each provider's standard first-party rate. A dash means the provider does not bill that item separately or no primary page states it.
What each costs in practice
The same three workloads every model page uses, worked from each model's standard list rates. Swap in your own token counts with the math on each model's page.
| Workload | MiniMax-M3 | Qwen3.8-Flash |
|---|---|---|
| One agent session1M input tokens and 100K output tokens, no caching. | $0.42 | $0.197 |
| A chat app10,000 requests a day at 2K input and 500 output tokens each. | $12.00$360.00 per 30-day month | $5.35$160.50 per 30-day month |
| The same session at a 90% cache hit90% of the input is read from the prompt cache. The one-time cache write premium is not included. | $0.204 | $0.0764 |
Show the math
MiniMax-M3
- One agent sessionInput: 1M x $0.30 = $0.30Output: 0.1M x $1.20 = $0.12
- A chat appInput: 20M x $0.30 = $6.00Output: 5M x $1.20 = $6.00
- The same session at a 90% cache hitCached input: 0.9M x $0.06 = $0.054Uncached input: 0.1M x $0.30 = $0.03Output: 0.1M x $1.20 = $0.12
Qwen3.8-Flash
- One agent sessionInput: 1M x $0.15 = $0.15Output: 0.1M x $0.47 = $0.047
- A chat appInput: 20M x $0.15 = $3.00Output: 5M x $0.47 = $2.35
- The same session at a 90% cache hitCached input: 0.9M x $0.016 = $0.0144Uncached input: 0.1M x $0.15 = $0.015Output: 0.1M x $0.47 = $0.047
Benchmarks under the same setting
No board in the benchmark directory lists both MiniMax-M3 and Qwen3.8-Flash under the same setting, so there is no like-for-like score to show here.
Set them up in your tools
Step-by-step guides whose official docs list MiniMax-M3 or Qwen3.8-Flash by model id.
- Use MiniMax in AiderLists MiniMax-M3
- Use MiniMax in Claude CodeLists MiniMax-M3
- Use MiniMax in ClineLists MiniMax-M3
- Use MiniMax in Codex CLILists MiniMax-M3
- Use MiniMax in CursorLists MiniMax-M3
- Use MiniMax in GitHub CopilotLists MiniMax-M3
- Use MiniMax in GooseLists MiniMax-M3
- Use MiniMax in Kilo CodeLists MiniMax-M3
- Use MiniMax in ZedLists MiniMax-M3
- Use Qwen in Claude CodeLists Qwen3.8-Flash
- Use Qwen in Codex CLILists Qwen3.8-Flash
- Use Qwen in Kilo CodeLists Qwen3.8-Flash
- Use Qwen in OpenCodeLists Qwen3.8-Flash
Model pages
FAQ
How do MiniMax-M3 and Qwen3.8-Flash compare on price?
On list price, MiniMax-M3 charges 2x Qwen3.8-Flash's input price ($0.30 vs $0.15 per 1M tokens), and MiniMax-M3 charges 2.6x Qwen3.8-Flash's output price ($1.20 vs $0.47 per 1M tokens). For one agent session of 1M input and 100K output tokens with no caching, that is $0.42 on MiniMax-M3 and $0.197 on Qwen3.8-Flash. Prices are each provider's first-party list rates, verified October 17, 2026 and October 17, 2026.
What are the context windows of MiniMax-M3 and Qwen3.8-Flash?
MiniMax-M3 and Qwen3.8-Flash both have a 1,000,000-token context window; max output is 524,288 and 131,072 tokens per response.
Are MiniMax-M3 and Qwen3.8-Flash open weights?
MiniMax-M3's weights are published at huggingface.co under the MiniMax Community License (Hugging Face license tag: other / minimax-community). Qwen3.8-Flash is closed-weights and is called through Alibaba Cloud (Qwen)'s API as qwen3.8-flash.
More comparisons
- Claude Haiku 5.5 vs MiniMax-M3Open weights vs closed
- Claude Haiku 5.5 vs Qwen3.8-FlashSimilar price, same tier
- GPT-6 Luna vs MiniMax-M3Open weights vs closed
- GPT-6 Luna vs Qwen3.8-FlashSimilar price, same tier
- DeepSeek V4.1 Flash vs MiniMax-M3Similar price, same tier
- DeepSeek V4.1 Flash vs Qwen3.8-FlashSimilar price, same tier
- GLM-5.3-Flash vs MiniMax-M3Similar price, same tier
- GLM-5.3-Flash vs Qwen3.8-FlashSimilar price, same tier
Sources
MiniMax-M3 verified from platform.minimax.io; Qwen3.8-Flash verified from alibabacloud.com. This page is as current as the older of the two: .
- https://platform.minimax.io/docs/guides/pricing
- https://www.alibabacloud.com/help/en/model-studio/model-pricing
- https://platform.minimax.io/docs/pricing/overview
- https://platform.minimax.io/docs/guides/text-generation
- https://platform.minimax.io/docs/api-reference/text-anthropic-api
- https://platform.minimax.io/docs/api-reference/text-chat-anthropic
- https://platform.minimax.io/docs/release-notes/models
- https://huggingface.co/MiniMaxAI/MiniMax-M3
- https://huggingface.co/MiniMaxAI/MiniMax-M3/raw/main/LICENSE
- https://modelstudio.console.alibabacloud.com/ap-southeast-1/model/market/detail/qwen3.8-flash
- https://www.qwencloud.com/models/qwen3.8-flash
- https://www.alibabacloud.com/help/en/model-studio/text-generation-model
- https://www.alibabacloud.com/help/en/model-studio/context-cache
- https://www.alibabacloud.com/help/en/model-studio/newly-released-models
Newsletter
Get model price changes as they ship
One email per week. New model releases, price cuts and what they mean for your bill.