Qwen3.8-Flash cache hits cost just $0.016 per million tokens
MaziyarPanahi · x · 2026-08-27
Alibaba's Qwen3.8-Flash on qwencloud is priced at $0.15/1M input tokens, $0.47/1M output tokens, and only $0.016/1M on cache hits. Developer MaziyarPanahi called the cache price crazy and said he now wants to see TTFT at 262K context.
Related event: Alibaba Releases Qwen3.8-Flash Preview with Major Efficiency Gains(22 posts)→
More from Infra
- llama.cpp PR adds dspark support for Nanbeige4.2-3B — jacek2023 · 2026-08-27
- Oak CEO: Every AI agent needs distinct identity and auditability — TechNadu · 2026-08-27
- Structured generation with 0% throughput overhead in production — remilouf · 2026-08-27
- M5 Pro vs M6 for local AI: memory bandwidth debate resurfaces — jonejy · 2026-08-27
- Self-hosted MiniMax H3 renders for $0.03, compared to fal's faster H3 Max — tobowers · 2026-08-27
- Volcengine Unveils Embodied AI Data Solution with Seedance Video Generation — 火山引擎 · 2026-08-27