Qwen 3.8 27B Hemingway
by NanoGPT · family qwen · listed Sep 2026RTVW
Qwen 3.8 27B Hemingway is an open-weight NVFP4 multimodal creative finetune for long-form prose, character dialogue, storytelling, and roleplay.
Current rates USD per 1M tokens
Input$0.25
Output$1.50
Cache read$0.125
Cache write—
Against the market
Input price vs. maker flagships ($/1M tokens)Output price vs. maker flagships ($/1M tokens)On the record
- Model id
qwen/qwen3.8-27b-hemmingway- Context window
- 262K tokens
- Max output
- 33K tokens
- Input modalities
- text, image
- Output modalities
- text
- Reasoning
- yes
- Tool calls
- yes
- Open weights
- yes
- Knowledge cutoff
- —
- Released
- 2026-09-21
- Last updated
- 2026-09-21
← All NanoGPT listings