Currently listed through these providers:
Model details
Qwen3.8 Max
Qwen3.8 Max is the flagship release in the Qwen3.8 series, positioned by its creators as the most capable member of the Qwen family to date and notable as the first Qwen-Max-class model whose weights will be open-sourced. It is built on the architectural foundation of Qwen 3.5 and scales to 2.4 trillion total parameters with 95 billion active, a Mixture-of-Experts style configuration that aims to deliver broad capability while keeping inference compute manageable. The release post frames the model as a multimodal reasoning system intended for complex reasoning, visual understanding, coding, and agentic workflows, marking a general-availability step beyond the earlier Qwen3.8 Max Preview.
The model's intended workload centers on long-horizon, end-to-end task completion rather than isolated prompt answering. According to the Qwen team's announcement, Qwen3.8 Max is designed to take multi-day coding and research projects from an empty folder to a finished deliverable with greater reliability, producing dependable results across coding, work, and research scenarios. OpenRouter's listing reinforces this practical profile by categorizing it for complex reasoning, coding, and agentic workflows, while reporting hosted availability through Alibaba Cloud at a price point of two dollars per million input tokens and six dollars per million output tokens, making it suitable for teams that need a frontier-class reasoning engine behind production pipelines and autonomous tool-using agents.
Quick Info
Powered by- Provider
- Vancine
- Model key
- qwen3.8-max
- Release date
- Aug 3, 2026
- Last updated
- Aug 3, 2026
- Input modalities
- Output modalities
- Capabilities
Cost
- Input token cost
- $1.60
- Output token cost
- $4.80
Limits
- Output tokens
- 131,072 tokens
- Context window
- 1,000,000 tokens
Latest news about Qwen3.8 Max
Videos about Qwen3.8 Max
More models around Qwen3.8 Max
This exact model name is also listed by 26 other providers.