6arXiv cs.CL (Computation and Language)·
PIVOT: Training-free sparse attention indexer cuts DeepSeek-V3.2 latency by up to 1.6x
PIVOT (Proxy Indexing Via One full-prefix Traversal) is a training-free drop-in replacement for the DeepSeek Sparse Attention (DSA) indexer that reduces the O(L²) per-query scan cost by grouping nearby queries and sharing a single prefix scan across the group. Two variants (PIVOT-Reuse and PIVOT-Refine) trade speed for fidelity, with PIVOT-Refine matching dense indexer accuracy. Evaluated on DeepSeek-V3.2 and GLM-5.1 across LongBench and RULER, PIVOT accelerates the indexer by up to 4x and reduces end-to-end latency by up to 1.6x at long context.
Related guides (3)
Related events (8)
DeepSeek Releases V3.2-Exp with Sparse Attention Architecture and 50%+ API Price Cut
DeepSeek has released DeepSeek-V3.2-Exp, an experimental model built on V3.1-Terminus that introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism designed to improve long-context performance and reduce compute costs during training and inference. Benchmarks indicate V3.2-Exp performs on par with V3.1-Terminus while achieving efficiency gains. The release is accompanied by a 50%+ API price reduction effective immediately, open-weights release on Hugging Face, a technical report, and GPU kernel code in TileLang and CUDA.
DeepSeek API Introduces Context Caching on Disk, Cutting Token Prices by ~90%
DeepSeek has launched a disk-based context caching service for its API, reducing cache-hit token pricing to $0.014 per million tokens versus $0.14 for cache misses—a 90% cost reduction. The system requires no code changes, runs automatically for prefix-matched inputs, and reduces first-token latency from ~13s to ~500ms on 128K prompts. DeepSeek attributes the feasibility of disk caching to the compact KV cache produced by its MLA (Multi-head Latent Attention) architecture in DeepSeek V2, which it claims makes it the first LLM API provider to deploy extensive disk caching at scale. The service supports up to 1 trillion tokens per day with no concurrency limits.
DeepSeek V4 Preview Release: 1.6T-param Pro and 284B Flash Models with 1M Context, Open-Sourced
DeepSeek has released DeepSeek-V4 as an open-weights preview, comprising two MoE variants: V4-Pro (1.6T total / 49B active parameters) and V4-Flash (284B total / 13B active parameters). Both models support 1M token context by default, enabled by a novel Token-wise compression and DeepSeek Sparse Attention (DSA) architecture. V4-Pro claims open-source SOTA on agentic coding benchmarks and world-class math/STEM/coding performance rivaling top closed-source models, while V4-Flash offers near-parity reasoning at lower cost and latency. The API is live today with OpenAI and Anthropic compatibility, and legacy model endpoints will be retired in July 2026.
DeepSeek-V3.1 Release: Hybrid Think/Non-Think Model with Agent-Focused Upgrades
DeepSeek has released V3.1, a hybrid inference model supporting both thinking and non-thinking modes in a single model, positioned as their first step toward the agent era. The model features improved tool use and multi-step agent task performance, with benchmarks showing gains on SWE-bench and Terminal-Bench, and faster thinking efficiency compared to DeepSeek-R1-0528. The base model received 840B tokens of continued pretraining for long-context extension, a new tokenizer, and open-source weights are available on HuggingFace. API updates include 128K context for both modes, Anthropic API format compatibility, and strict function calling support in beta.
DeepSeek-V3: 671B MoE Open-Source Model with 3x Speed Improvement
DeepSeek releases V3, a 671B parameter Mixture-of-Experts model with 37B activated parameters, trained on 14.8T tokens. The model runs at 60 tokens/second (3x faster than V2) and is fully open-source with weights and paper released. API pricing is set at $0.27/M input tokens and $1.10/M output tokens starting February 8, positioning it as a low-cost frontier alternative. DeepSeek signals future multimodal capabilities in the ecosystem.
SLAI T-Rex: Full-parameter post-training of DeepSeek-V4 family on Ascend NPU SuperPOD achieves 34% MFU
Researchers present SLAI T-Rex, an end-to-end optimization framework for full-parameter post-training of trillion-parameter MoE models on Huawei Ascend NPU SuperPOD infrastructure, using the DeepSeek-V4 model family as the target workload. The system achieves 34.22% Model FLOPs Utilization, a 2.93x improvement over the open-source baseline, through hierarchical optimizations spanning model parallelism, communication orchestration, and kernel execution. Building on this infrastructure, the team develops a domain-specialized CPT and SFT pipeline for Operations Research tasks using DeepSeek-V4-Flash, producing a model that achieves 71.81% zero-shot Pass@1 on OR benchmarks, outperforming GPT-5.4-Mini by ~4 percentage points. The work is notable both as a non-GPU large-scale training system report and as a demonstration of domain specialization for complex mathematical reasoning.
DeepSeek Reasonix: Native Coding Agent with High Caching and Low Cost
DeepSeek Reasonix is a coding agent built natively on DeepSeek models, emphasizing high prompt caching rates and low inference cost. The project attracted significant Hacker News engagement (349 points, 171 comments), suggesting community interest in cost-efficient agentic coding workflows. It appears to be an open-source or community-developed tool rather than an official DeepSeek Labs release.
DeepSeek open-sources DSpark speculative decoding module, achieving 50%+ throughput gains in production
Researchers from Peking University and DeepSeek introduced DSpark, a speculative decoding module that dynamically adjusts verification depth based on server load, achieving 57–85% faster per-user token generation and 51–52% higher total throughput compared to DeepSeek's previous production drafter. The team released checkpoints DeepSeek-V4-Pro-DSpark and DeepSeek-V4-Flash-DSpark on Hugging Face under an MIT license, with the draft module attaching to frozen target model weights. Key innovations include a parallel drafting backbone (adapted from DFlash), a Markov head for sequential token coherence correction, a calibrated confidence head, and a load-aware scheduler that trades draft length against server capacity. Results generalize across model families including Qwen3 and Gemma4.