--- title: DeepSeek V4.1 Flash's 4x KV cache shrink is the agentic... description: DeepSeek released V4.1 Flash, an open model with 500B+ parameters featuring a 4x reduction in KV cache size compared to V4-Flash, down to roughly 890 bytes per... canonical: https://daily.dev/posts/deepseek-v4-1-flash-s-4x-kv-cache-shrink-is-the-agentic-infrastructure-play-nobody-expected-2xbydasia twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: DeepSeek V4.1 Flash's 4x KV cache shrink is the agentic infrastructure play nobody expected | daily.dev og:description: DeepSeek released V4.1 Flash, an open model with 500B+ parameters featuring a 4x reduction in KV cache size compared to V4-Flash, down to roughly 890 bytes per... og:url: https://daily.dev/posts/deepseek-v4-1-flash-s-4x-kv-cache-shrink-is-the-agentic-infrastructure-play-nobody-expected-2xbydasia og:image: https://api.daily.dev/og/posts/2xbydASIa.png og:image:alt: DeepSeek V4.1 Flash's 4x KV cache shrink is the agentic infrastructure play nobody expected og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # DeepSeek V4.1 Flash's 4x KV cache shrink is the agentic infrastructure play nobody expected **[Trends](https://daily.dev/sources/trends)** · 2 min read · 3 upvotes · 0 comments ## Summary DeepSeek released V4.1 Flash, an open model with 500B+ parameters featuring a 4x reduction in KV cache size compared to V4-Flash, down to roughly 890 bytes per token. The gains come from a Causal Encoder-Decoder architecture that splits a 40-layer model into separate prefill and decode halves, combined with cross-layer KV sharing and 4-bit cache storage. At maximum reasoning effort, the model reportedly beats the larger V4-Pro on agentic benchmarks like DeepSWE v1.1, Terminal-Bench 2.1, and AutomationBench, though it consumes significantly more tokens during reasoning, making it better suited for long agentic workloads than latency-sensitive use cases. ## Content DeepSeek dropped V4.1 Flash and the headline isn't the benchmark numbers — it's the architecture. The model introduces something called a Causal Encoder-Decoder (CED) design, which splits the 40-layer transformer into two distinct halves: 20 layers for reading (prefill, 8B active parameters) and 20 layers for writing (decode, 16B active parameters). Standard transformers treat both phases identically. CED doesn't. The decoder pulls its KV cache directly from encoder outputs instead of recomputing per layer, and combined with cross-layer KV sharing and 4-bit cache storage, the result is a global KV cache of roughly 890 bytes per token — about a quarter of V4-Flash's footprint, and apparently 437x smaller than V1. That's not a rounding error. That's a different infrastructure conversation. The benchmark story is also genuinely interesting: at maximum reasoning effort, V4.1 Flash outperforms the larger V4-Pro on agentic benchmarks including DeepSWE v1.1, Terminal-Bench 2.1, and AutomationBench. The gains are biggest on long multi-step tasks, which is exactly where KV cache size and input-scaling compute actually hurt you in production. DeepSeek's argument is essentially that for agentic workloads, infrastructure cost matters more than leaderboard position — and they built the model to prove it. The tradeoffs are real. The model burns tokens hard during reasoning, and 500B+ parameters isn't small. But the size-to-capability ratio is drawing attention from people who care about what it costs to actually run these things, not just what they score. For context on why the cache reduction matters so much: speculative decoding research (a separate but related thread in the inference optimization space) shows that methods like EAGLE and LayerSkip get 1.8x–3.6x speedups partly by reducing redundant computation across layers. CED is attacking a similar problem from the architecture level rather than the serving stack level. Whether that's the right place to solve it is an open question, but the numbers suggest it's at least a valid one. ## Questions this post answers ### What is the KV cache size of DeepSeek V4.1 Flash compared to V4-Flash? DeepSeek V4.1 Flash reduces KV cache size by 4x compared to V4-Flash, down to roughly 890 bytes per token. This comes from a Causal Encoder-Decoder architecture that splits the 40-layer model into 20 prefill layers and 20 decode layers, plus cross-layer KV sharing and 4-bit cache storage, cutting memory footprint to about a quarter of V4-Flash's. _Teams sizing inference infrastructure for agentic workloads can follow model efficiency shifts like this on daily.dev._ ### How does the Causal Encoder-Decoder architecture in DeepSeek V4.1 Flash work? It splits a 40-layer transformer into two halves: 20 layers handle prefill with 8B active parameters per token, and 20 layers handle decode with 16B active parameters per token. The decoder pulls its KV cache directly from encoder outputs rather than recomputing per layer, reducing redundant compute and memory overhead compared to standard transformers that use the same layers for both reading and generating. _Developers evaluating new model architectures for agent pipelines can track these design shifts via daily.dev._ ### Does DeepSeek V4.1 Flash outperform V4-Pro on agentic tasks? At maximum reasoning effort, DeepSeek V4.1 Flash reportedly beats the larger V4-Pro model on agentic benchmarks including DeepSWE v1.1, Terminal-Bench 2.1, and AutomationBench. The tradeoff is that it burns significantly more tokens during reasoning, which suits long agentic workloads where compute is already the cost driver but is less favorable for latency-sensitive or token-budget-constrained use cases. _Anyone choosing between agentic model sizes can weigh these tradeoffs using daily.dev._ ## Similar posts on daily.dev - [DeepSeek-V4: a million-token context that agents can actually use](https://daily.dev/posts/deepseek-v4-a-million-token-context-that-agents-can-actually-use-uczd4kivx) · Hugging Face · 27 upvotes · 2 comments --- Tags: [#deep-learning](https://daily.dev/tags/deep-learning), [#agentic-ai](https://daily.dev/tags/agentic-ai), [#ai-inference](https://daily.dev/tags/ai-inference), [#deepseek](https://daily.dev/tags/deepseek) [View this post on daily.dev](https://daily.dev/posts/deepseek-v4-1-flash-s-4x-kv-cache-shrink-is-the-agentic-infrastructure-play-nobody-expected-2xbydasia) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"DeepSeek V4.1 Flash's 4x KV cache shrink is the agentic infrastructure play nobody expected","url":"https://daily.dev/posts/deepseek-v4-1-flash-s-4x-kv-cache-shrink-is-the-agentic-infrastructure-play-nobody-expected-2xbydasia","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/deepseek-v4-1-flash-s-4x-kv-cache-shrink-is-the-agentic-infrastructure-play-nobody-expected-2xbydasia"},"datePublished":"2026-09-18T17:22:08.972Z","dateModified":"2026-09-19T05:31:12.598Z","description":"DeepSeek released V4.1 Flash, an open model with 500B+ parameters featuring a 4x reduction in KV cache size compared to V4-Flash, down to roughly 890 bytes per...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/33427aa5bcbbab36a0d2bad86c78617f?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/33427aa5bcbbab36a0d2bad86c78617f?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Trends","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Trends","logo":"https://media.daily.dev/image/upload/s--ZfSp3asX--/f_auto,q_auto/v1780996004/logos/trends?_a=BAMAMiWQ0","url":"https://daily.dev/sources/trends"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/deepseek-v4-1-flash-s-4x-kv-cache-shrink-is-the-agentic-infrastructure-play-nobody-expected-2xbydasia","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":3},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"deep-learning,agentic-ai,ai-inference,deepseek","timeRequired":"PT2M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Trends","item":"https://daily.dev/sources/trends"},{"@type":"ListItem","position":3,"name":"DeepSeek V4.1 Flash's 4x KV cache shrink is the agentic infrastructure play nobody expected"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/deepseek-v4-1-flash-s-4x-kv-cache-shrink-is-the-agentic-infrastructure-play-nobody-expected-2xbydasia#faq","mainEntity":[{"@type":"Question","name":"What is the KV cache size of DeepSeek V4.1 Flash compared to V4-Flash?","acceptedAnswer":{"@type":"Answer","text":"DeepSeek V4.1 Flash reduces KV cache size by 4x compared to V4-Flash, down to roughly 890 bytes per token. This comes from a Causal Encoder-Decoder architecture that splits the 40-layer model into 20 prefill layers and 20 decode layers, plus cross-layer KV sharing and 4-bit cache storage, cutting memory footprint to about a quarter of V4-Flash's. Teams sizing inference infrastructure for agentic workloads can follow model efficiency shifts like this on daily.dev."}},{"@type":"Question","name":"How does the Causal Encoder-Decoder architecture in DeepSeek V4.1 Flash work?","acceptedAnswer":{"@type":"Answer","text":"It splits a 40-layer transformer into two halves: 20 layers handle prefill with 8B active parameters per token, and 20 layers handle decode with 16B active parameters per token. The decoder pulls its KV cache directly from encoder outputs rather than recomputing per layer, reducing redundant compute and memory overhead compared to standard transformers that use the same layers for both reading and generating. Developers evaluating new model architectures for agent pipelines can track these design shifts via daily.dev."}},{"@type":"Question","name":"Does DeepSeek V4.1 Flash outperform V4-Pro on agentic tasks?","acceptedAnswer":{"@type":"Answer","text":"At maximum reasoning effort, DeepSeek V4.1 Flash reportedly beats the larger V4-Pro model on agentic benchmarks including DeepSWE v1.1, Terminal-Bench 2.1, and AutomationBench. The tradeoff is that it burns significantly more tokens during reasoning, which suits long agentic workloads where compute is already the cost driver but is less favorable for latency-sensitive or token-budget-constrained use cases. Anyone choosing between agentic model sizes can weigh these tradeoffs using daily.dev."}}]} ```