DeepSeek V4.1 Flash, served from DeepSeek's official DeepSeek-V4.1-Flash-1017 release: the October 17 update of DeepSeek's latest flash model, a 552B mixture-of-experts (8B active per token on prefill, 16B on decode) built for fast agentic coding, with native image understanding and a 1M-token context window. Updated on October 20, 2026 from the original DeepSeek-V4.1-Flash weights to 1017, which DeepSeek reports as stronger on reasoning, coding, agentic and visual benchmarks, with shorter outputs. This id follows future V4.1 Flash updates; umans-deepseek-v4.1-flash-1017 stays on 1017. Reasoning has four modes: non-think (none), think low (low), think high (high, the default) and think max (max). Billed per token ($0.15 / $0.60 / $0.028 per 1M; input / output / cache read). Served on our own GPU infrastructure with high availability.
Umans DeepSeek V4.1 Flash
also served as umans-deepseek-v4.1-flash-1017
728.1tok/s
throughput · p50 · last 5 min
97ms
TTFT · p50 · last 5 min
99.91%
uptime · 24h
updated to 1017 on Oct 20, 2026
Context
1049K
Max output
393K
Recommended
393K
Vision
Yes
Tools
Yes
Reasoning
Toggle · none/low/high/max
Weights
Trends
90 days agopre-release before Sep 18, 2026today
updated to 1017 on Oct 20, 2026 · dotted: 1017 in Labs, before the update
90 days agopre-release before Sep 18, 2026today
updated to 1017 on Oct 20, 2026 · dotted: 1017 in Labs, before the update
Changelog
Oct 202026
Updated: Umans DeepSeek V4.1 Flash now serves 1017 Updated
umans-deepseek-v4.1-flash now serves DeepSeek-V4.1-Flash-1017, DeepSeek's October 17 update to V4.1 Flash, which ran here as a Labs experiment from October 17: the same 552B mixture-of-experts (8B active per token on prefill, 16B on decode), native image understanding and 1M context window, on new weights that DeepSeek reports as stronger on reasoning, coding, agentic and visual benchmarks, with shorter outputs. Same price: $0.15 / $0.60 / $0.028 per 1M (input / output / cache read). This id follows future V4.1 Flash updates; to stay on this exact checkpoint, use umans-deepseek-v4.1-flash-1017. The status page marks the update and shows the lab's points next to the production line.
Sep 182026
Released pay-per-token: Umans DeepSeek V4.1 Flash Released
umans-deepseek-v4.1-flash joins the lineup: DeepSeek-V4.1-Flash, the checkpoint that served here as the seat-gated pre-release lab since September 10 - DeepSeek's latest flash model, a 552B mixture-of-experts (8B active per token on prefill, 16B on decode) built for fast agentic coding, with native image understanding and a 1M context window. Reasoning has four modes: non-think, think low, think high (the default) and think max. Billed per token: $0.15 / $0.60 / $0.028 per 1M (input / output / cache read). The pre-release window's metrics stay on the model's status page as its pre-release period (before Sep 18). The free Labs experiment keeps running on umans-deepseek-v4.1-flash-lab for what comes next.