Back to Models
Qwen: Qwen3.8 2.4T A95B
Qwentext9.1 / 10 Overall Rating
Qwen3.8 2.4T A95B is an open-weight sparse mixture-of-experts model created by Alibaba's Qwen team, utilizing 95 billion active parameters out of 2.4 trillion total. It provides a 1,048,576 token context window and a 131,072 max output length at pricing lower than GPT-4o and Claude 3.5 Sonnet.
Context Window
1049K
Knowledge Cutoff
2025
Max Output
16K
License
Proprietary
Where it Excels (Pros)
- Massive 1M token context window
- 131K max output token limit
- Cheaper than GPT-4o and Sonnet
Limitations (Cons)
- High hosting hardware requirements
- Variable latency from MoE routing
Benchmark Breakdown
Reasoning (MMLU)82%
Coding (HumanEval)78%
Mathematics (MATH)76%
Long-Context Retrieval95%
Pricing Matrix
Input cost / 1M tokens$2.00
Output cost / 1M tokens$6.00
Compare Qwen: Qwen3.8 2.4T A95B with other frontier models: