Ollama launches off-peak pricing: DeepSeek-V4 models half price nights and weekends

ollama · x · 2026-09-06

Ollama officially introduced off-peak token rates for its cloud inference: DeepSeek-V4-Flash and Pro are half price outside 12:00–18:00 UTC on weekdays (5–11am Pacific) and all day on weekends. More models will get off-peak pricing soon.

Regular per-million-token prices: deepseek-v4-flash at $0.22 input / $0.66 output, deepseek-v4-pro at $0.66 / $1.98, alongside gemma4, glm-5.x, kimi-k3, minimax and qwen3.5:397b. DeepSeek models are hosted in the US and Europe with zero data retention (ZDR). Subscription tiers include Free, Pro ($20/mo), Max ($100/mo) and Team ($500/mo).

Related event: Ollama Cloud Cuts DeepSeek-V4 Token Prices in Half Off-Peak(2 posts)→

Original post →

More from Infra

Infra channel →