Qwen3.8 2.4T A95B
by Charm Hyper · family qwen · listed Aug 2026RTW
Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows
Current rates USD per 1M tokens
Input$2.00
Output$6.00
Cache read$0.25
Cache write—
Against the market
On the record
- Model id
qwen3.8-2.4t-a95b- Context window
- 1M tokens
- Max output
- 128K tokens
- Input modalities
- text
- Output modalities
- text
- Reasoning
- yes
- Tool calls
- yes
- Open weights
- yes
- Knowledge cutoff
- —
- Released
- 2026-08-26
- Last updated
- 2026-08-26
- Provider docs
- hyper.charm.land
Back of the envelope
≈
Also listed at
- NanoGPT$2.00 in · $6.00 out
- AIHubMix$2.00 in · $6.00 out
- Requesty$2.00 in · $6.00 out
- Vercel AI Gateway$2.00 in · $6.00 out
- Modal$2.00 in · $6.00 out
- Deep Infra$2.00 in · $6.00 out
- Kilo Gateway$2.00 in · $6.00 out
- OpenRouter$2.00 in · $6.00 out
- Eden AI$2.00 in · $6.00 out
- Hugging Face$2.50 in · $6.25 out
- Merge Gateway$2.50 in · $6.25 out
- Cortecs$2.50 in · $6.00 out