DeepSeek launches V4.1-Flash: 552B backbone, 1M-token context

DeepSeek released DeepSeek-V4.1-Flash around September 11, the smallest member of its new architecture family, updating both its web app and mobile app with a "fast, expert, visual" positioning. It is a multimodal MoE with a 552B backbone (763B total parameters per Ollama's listing), native visual understanding, up to 1M-token context, and support for vision, tools, and thinking.

Confirmed

DeepSeek officially confirmed V4.1-Flash is the smallest model in the new family, promising higher capability, faster inference, and higher throughput, paving the way for larger models.
The model uses a Causal Encoder-Decoder architecture with 4x smaller KV cache.
Fireworks launched the model targeting coding and agent use cases; it is also available on the FLock API platform and Ollama (cloud mode).
Per APPSO, the backbone doubled from 284B to 552B while running faster: measured at 420–507 tok/s, with one user reporting 217 tok/s.

Unconfirmed

Parameter figures differ across sources: reposts cite 552B, while Ollama lists 763B total / 552B backbone; the official figure needs verification.
Claims such as "website-building cost 1/40 of Claude" and "Agent scores surpassing Claude" come from reposts and media, with unclear test conditions.

Why it matters

As the first small-size member of DeepSeek's new architecture family, its capability and cost profile sets the tone for larger models to come.
A 552B backbone combined with 1M-token context and 4x KV-cache compression poses direct competitive pressure on closed-source peers in long-context and agentic scenarios.

2026-09-11 ~ 2026-09-12 · 5 related posts

Full story(18 episodes)→

Episode 1: DeepSeek V4.1 Flash Opens Limited Internal Beta with New Architecture and Native Multimodality(2026-09-08, 18 posts)
Episode 2: DeepSeek V4.1 Flash Tested: Blazing 350 Tokens/s but Still Experimental(2026-09-08, 2 posts)
Episode 3: DeepSeek cuts V4-Flash API prices with new peak/off-peak billing from Sept 10(2026-09-08, 6 posts)
Episode 4: DeepSeek V4.1 Flash Nears Frontier Models at a Fraction of the Cost, Independent Tests Show(2026-09-09, 15 posts)
Episode 5: DeepSeek Releases Open-Source V4.1 Flash, Beating Its Own Flagship at a Fraction of the Cost(2026-09-09, 57 posts)
Episode 6: DeepSeek V4.1 Flash Leaks: 552B Asymmetric MoE Claimed to Rival GPT-5.6 Sol(2026-09-10, 21 posts)
Episode 7: DeepSeek V4.1 Flash Benchmark Results Leak, Sparking Community Discussion(2026-09-10, 2 posts)
Episode 8: DeepSeek Shrinks Per-Token KV Cache 54x in Nine Months to Sub-KB Levels(2026-09-10, 6 posts)
Episode 9: Leaked DeepSeek-V4.1-Flash Touted as New Open-Source King(2026-09-10, 3 posts)
Episode 10: Bug Hunt Bench Ranks Frontier Coding Models on 105 Real Bugs(2026-09-10, 13 posts)
Episode 11: DeepSeek V4.1 Flash Confirmed to Center on YOCO Architecture(2026-09-10, 7 posts)
Episode 12: Leaked DeepSeek V4.1 Flash Tech Report Reveals KV Cache-Obsessed Flagship(2026-09-11, 44 posts)
Episode 13: DeepSeek V4.1 Flash tops Vals open-source index at $0.30 per run(2026-09-11, 2 posts)
Episode 14: DeepSeek launches V4.1-Flash: 552B backbone, 1M-token context(2026-09-11, 5 posts)
Episode 15: DeepSeek V4.1 Flash scores 40 in AA index despite rising hallucinations(2026-09-11, 2 posts)
Episode 16: Inside KV-Cache Sharing: How DeepSeek CED and GLM 5.2 Differ(2026-09-11, 4 posts)
Episode 17: DeepSeek 4.1 Flash Shows Big Gains in Kernel Engineering and Cost(2026-09-11, 2 posts)
Episode 18: SGLang delivers Day 0 support for DeepSeek V4.1 Flash, hitting 873 tok/s(2026-09-12, 2 posts)

Primary sources

DeepSeek-V4.1-Flash hits Ollama: 552B MoE backbone with 1M context via KV cache compression — ollama · 2026-09-11 07:14 (UTC+08:00)
DeepSeek launches V4.1-Flash with 1M-token context and 4x smaller KV-cache — matlabulous · 2026-09-11 15:40 (UTC+08:00)
DeepSeek V4.1 Flash: 552B params at 420-507 tok/s, matches Claude Opus-5 on agent benchmarks — APPSO · 2026-09-12 11:00 (UTC+08:00)

DeepSeek-V4.1-Flash lands on Fireworks: 552B MoE for coding and agents at 1/40th claimed cost — lqiao · 2026-09-11
[source] DeepSeek-V4.1-Flash hits Ollama: 552B MoE backbone with 1M context via KV cache compression — ollama · 2026-09-11
[source] DeepSeek launches V4.1-Flash with 1M-token context and 4x smaller KV-cache — matlabulous · 2026-09-11
DeepSeek ships V4.1-Flash with native vision; builds landing pages for 1/40th of Claude's cost — togethercompute · 2026-09-12
[source] DeepSeek V4.1 Flash: 552B params at 420-507 tok/s, matches Claude Opus-5 on agent benchmarks — APPSO · 2026-09-12