DeepSeek released V4.1-Flash, the smallest model of a new architecture family with native multimodal vision: a 552B backbone (763B total safetensors params including a 196B Engram conditional-memory module), 8B activated per token during prefill and 16B during decode, 1 shared + 384 routed experts with 6 active. The CED design organizes 40 layers as a 20-layer causal encoder followed by a 20-layer decoder whose KV cache is projected from final encoder hidden states; CSA2 provides three static attention modes (Full/Reindex/Reuse) with a hierarchical sparse indexer, FP4 (E2M1) main KV caching at 890 bytes/token (~1/4 of V4-Flash; SWA bounded replay cuts persistent KV to ~1/8), plus Single-Pass mHC kernels and DSpark speculative decoding. Pre-trained on 45T multimodal tokens; context extended to 1M. Benchmarks (reasoning_effort=100): Terminal-Bench 2.1 90.6, TB 3.0 30.0, TB 4.0 31.2, DeepSWE v1.1 74.2 (vs V4-Flash 54.4), GPQA Diamond 90.9, Codeforces 3471, HLE-with-tools 63.9, CyberGym 88.1, ExploitGym 15.3. MIT license, ungated, day-0 vLLM/SGLang/Docker Model Runner/Transformers support (deepseek-recipe libraries; no Jinja chat template). API: the model name is now deepseek-flash ($0.30 peak / $0.15 off-peak input, $1.20 / $0.60 output per 1M tokens; cache-hit $0.006 peak); V4-Flash and V4-Flash-Vision-Exp are retired from the API with legacy names temporarily routed to V4.1-Flash; V4 Pro API service continues after 2026-09-14 with unchanged billing; reasoning effort is a controllable integer 1-100. 75,774 downloads + 1,741 likes within ~1 day; #1 on HF trending at capture.
DeepSeek V4.1-Flash: new CED/CSA2 architecture with 1M context, open weights
DeepSeek V4.1-Flash is an open-weight multimodal model with 1M context and a sparse CED/CSA2 architecture. The October 17 1017 update replaced the website, app and deepseek-flash API alias without changing price or interface and shipped a new MIT checkpoint. Artificial Analysis now independently scores the hosted Max configuration at 45 on its Intelligence Index, $0.26 per task, 224.7 output tokens per second and 1.12-second TTFT; the run used 210M output tokens, so high verbosity remains a cost caveat.
Imported from the existing research archive. Migration preserves the recorded judgment and is not a fresh verification.
What this means for you
For teams choosing an open-weight coding and agent model, independent API results now support the claim that 1017 is fast and competitively priced, while also exposing its unusually high token use. Trial decisions should measure total task cost and success on your own workload; the independent run does not reproduce the downloadable weights.
Evidence & related material
DeepSeek released the 1017 weights and updated its hosted alias to the new checkpoint.
Artificial Analysis independently measured the hosted DeepSeek V4.1 Flash 1017 Max configuration for index score, speed, latency and task cost, while showing high output-token use.
NVIDIA published an NVFP4 checkpoint of DeepSeek V4.1 Flash 1017 for Blackwell-class GPUs.
Subsequent changes
V4.1-Flash-1017 weights and hosted update go live
Artificial Analysis independently measures 1017 speed, task cost and verbosity
NVIDIA publishes a 1017 NVFP4 checkpoint for Blackwell GPUs
Technical details & scores
| Release | HF repo created 2026-09-10T02:17:58Z; MIT; image-text-to-text; 75,774 dl + 1,741 likes in ~1 day (captured 2026-09-11T18:34Z) |
|---|---|
| Architecture | 552B backbone / 763B total safetensors (incl. 196B Engram conditional memory); 8B prefill / 16B decode activated; MoE 1 shared + 384 routed, 6 active; CED = 20-layer causal encoder + 20-layer decoder (KV projected from encoder final hidden states); CSA2 static attention modes Full/Reindex/Reuse + hierarchical sparse indexer; FP4 E2M1 main KV, 890 B/token (~1/4 V4-Flash; ~1/8 with SWA bounded replay); Single-Pass mHC; DSpark speculative decoding; DeepSeek-ViT vision (2D-RoPE, 3x3 pixel-unshuffle) |
| Training | 45T multimodal tokens pre-training; sparse attention at 64K; context extended to 1M at 34T; SFT -> RL -> on-policy distillation; reasoning_effort integer 1-100 |
| Benchmarks | TB 2.1 90.6 / TB 3.0 30.0 / TB 4.0 31.2; DeepSWE v1.1 74.2 (V4-Flash 54.4); GPQA Diamond 90.9; Codeforces 3471; HLE 36.8 (39.1 text-only); HLE w/tools 63.9; MathArena Apex 65.6; CyberGym 88.1; SEC-Bench Pro 62.8; ExploitGym 15.3; MMMU-Pro 56.5; DocVQA 95.6 |
| API | model name deepseek-flash; input $0.30 peak / $0.15 off-peak, output $1.20 / $0.60 per 1M; cache-hit $0.006 peak; concurrency 2500; max output 384K; V4-Flash + V4-Flash-Vision-Exp retired (legacy names temporarily routed); V4 Pro continues past 2026-09-14 unchanged billing |
| Serving | day-0 vLLM / SGLang / Docker Model Runner / Transformers; deepseek-recipe Rust libs + Python reference; no Jinja chat template |
Historical update notes
· 2026-09-13 run, serving-ecosystem check: vLLM now carries 63 issues/PRs mentioning V4.1-Flash (13 on 2026-09-12 alone - encoder CUDA-graph support #56625 17:54Z, mHC post-block folding into the delayed pre-projection #56633, ROCm gfx950 re-enable #56638; ROCm performance RFC #56506 opened 9/11); SGLang shows 32 matches (MI350X/gfx950 support #39186, Mooncake Engram KV backend with CUDA Graph #39205, prefill context parallelism #39179) plus a third-party report that fp8 wo_a absorb GEMM output is silently ~25% wrong when DEEPGEMM_SCALE_UE8M0=false on SM121 (#39193, 06:21Z). The HF model card still ships no serving recipe (community request in #13); deepseek-ai/deepseek-recipe is a prompt-format conversion library, not serving. 15+ third-party deployment repos created/updated on 9/12 target DGX Spark/GB10, Mac Studio 512GB and EXL3 quantizations; LiteLLM pricing PRs #40919/#40920 (9/12 23:36Z) add V4.1-Flash across five providers. Downloads @2026-09-13T00:03Z: 140,636 (+85.6% vs ~75.8k @9/11, ~2.5 days after release) + 2,005 likes. No third-party Terminal-Bench/SWE run yet. Stays TRIAL.
· 2026-09-14 report run: vLLM main merged DeepSelect sparse-indexer TopK (#56464; 191 us vs 616 us on GB200 at batch 256 / 1M KV, kernel-only) and production paths for Engram DP sharding, async CPU-offload prefetch and shared-memory tables (#56512; two-node TP1xDP8 validation, <=0.9% reported TTFT gain). Not yet tagged. Recommendation stays TRIAL.