model releases

🔹 Amid recent attention, a quick reminder: please rely only on our official accounts for DeepSeek news. Statements from other channels do not reflect our views. 🔹 Thank you for your continued trust. W

An update on recent Claude Code quality reports It turns out the high volume of complaints that Claude Code was providing worse quality results over the past two months was grounded in real problems.

DeepSeek released a new open-weights model with strong benchmark performance, though Mollick notes that benchmark results may not fully capture the capabilities of open models compared to closed syste

... and there's me thinking that sending it out at 9pm Pacific Time on a Thursday evening was safe, surely there wouldn't be any news this evening that I might want to include in there... https://x.co

Anthropic: Anthropic details Project Deal, a marketplace experiment where Claude models bought, sold, and negotiated personal belongings on behalf of Anthropic employees — At Anthropic, we're interest

arXiv:2505.09971v3 Announce Type: replace Abstract: Airborne laser scanning (ALS) point cloud semantic segmentation is a fundamental task for large-scale 3D scene understanding. Fixed models deployed

API is Available Today! 🔹 Keep base_url, just update model to deepseek-v4-pro or deepseek-v4-flash. 🔹 Supports OpenAI ChatCompletions & Anthropic APIs. 🔹 Both models support 1M context & dual modes (T

arXiv:2604.21199v1 Announce Type: cross Abstract: Time series question-answering (TSQA), in which we ask natural language questions to infer and reason about properties of time series, is a promising

As a point of comparison, our commercial document OCR solution LlamaParse wins on all relevant dimensions (except content faithfulness against Opus) and is 1.25c / page prior to an enterprise plan. Co

arXiv:2604.21766v1 Announce Type: new Abstract: Existing audio question answering benchmarks largely emphasize sound event classification or caption-grounded queries, often enabling models to succeed

arXiv:2601.10863v3 Announce Type: replace Abstract: Traditional time series forecasting methods optimize for accuracy alone. This objective neglects temporal consistency, in other words, how consisten

arXiv:2604.21724v1 Announce Type: new Abstract: Large token-indexed lookup tables provide a compute-decoupled scaling path, but their practical gains are often limited by poor parameter efficiency and

arXiv:2604.21134v1 Announce Type: new Abstract: Vision-Language Models (VLMs) frequently misread values, hallucinate details, and confuse overlapping elements in charts. Current approaches rely solely

arXiv:2604.21344v1 Announce Type: cross Abstract: Charts are widely used to present complex information. Deriving meaningful insights in real-world contexts often requires interpreting multiple relate

arXiv:2604.21508v1 Announce Type: new Abstract: Protein-ligand bioactivity data published in the literature are essential for drug discovery, yet manual curation struggles to keep pace with rapidly gr

arXiv:2604.20945v1 Announce Type: cross Abstract: Effective safety auditing of large language models (LLMs) demands tools that go beyond black-box probing and systematically uncover vulnerabilities ro

🚨BREAKING: Hugging Face just open-sourced an AI intern that reads ML papers, trains models, and ships the final model for you. It’s called ML Intern. And this is not another AI coding demo that prints

Developers can build with DeepSeek V4 through NVIDIA GPU-accelerated endpoints on build.nvidia.com, with hosted endpoints providing a fast way to prototype before moving to self-hosted deployment. Dee

arXiv:2604.21718v1 Announce Type: cross Abstract: Video-language models (VLMs) learn to reason about the dynamic visual world through natural language. We introduce a suite of open datasets, benchmark

arXiv:2604.21277v1 Announce Type: new Abstract: We introduce MMTR-Bench, a benchmark designed to evaluate the intrinsic ability of Multimodal Large Language Models (MLLMs) to reconstruct masked text d

arXiv:2604.21251v1 Announce Type: cross Abstract: Large language models (LLMs) trained on unfiltered corpora inherently risk retaining sensitive information, necessitating selective knowledge unlearni

arXiv:2604.20845v1 Announce Type: cross Abstract: Next Point-of-Interest (POI) recommendation plays a crucial role in location-based services by predicting users' future mobility patterns. Existing me

Chinese AI company DeepSeek released a preview of its hotly anticipated next-generation AI model V4 on Friday, saying that the open-source model can compete with leading closed-source systems from US

Ben Jiang / South China Morning Post: China's top market regulator says it is launching a six-month crackdown on the country's online ad sector, targeting malpractices including the misuse of AI — Mar

arXiv:2604.21308v1 Announce Type: cross Abstract: Enterprise LLM agents can dramatically improve workplace productivity, but their core capability, retrieving and using internal context to act on a us

Cloud-hosted models will become more niche/seldom-used as much cheaper local models fulfill that which 90% of consumers need from AI. Lots of models to buy/use, but very few hardware options upon whic

arXiv:2604.21270v1 Announce Type: cross Abstract: There has been remarkable progress over the past decade in establishing finite-sample, non-asymptotic bounds on recovering unknown system parameters f

arXiv:2603.18677v2 Announce Type: replace-cross Abstract: Artificial intelligence is increasingly embedded in human decision making. In some cases, it enhances human reasoning. In others, it fosters e

This post from Swyx appears to be a humorous or provocative invitation directed at Anthropic AI, suggesting a particular country has the highest per-capita adoption or focus on AI technology and encou

arXiv:2604.21789v1 Announce Type: cross Abstract: Competing firms that serve shared customer populations face a fundamental information aggregation problem: each firm holds fragmented signals about ri

arXiv:2512.16001v2 Announce Type: replace-cross Abstract: Measuring the statistical dependence between observed signals is a primary tool for scientific discovery. However, biological systems often ex

Congrats to @deepseek_ai team! Doing the numbers I would estimate: Pro < 14m for the final training run Flash < 4m Ratio of active params x total training tokens vs v3 Total compute costs (data prep,

arXiv:2509.21361v2 Announce Type: replace-cross Abstract: Large language model (LLM) providers boast big numbers for maximum context window sizes. To test the real world use of context windows, we 1)

arXiv:2604.21241v1 Announce Type: cross Abstract: Vision--Language--Action (VLA) models often use intermediate representations to connect multimodal inputs with continuous control, yet spatial guidanc

arXiv:2506.21546v4 Announce Type: replace-cross Abstract: Segmentation Vision-Language Models (VLMs) have significantly advanced grounded visual understanding, yet they remain prone to pixel-grounding

arXiv:2604.21286v1 Announce Type: cross Abstract: Cacioli (2026) showed that the K-way energy probe on standard discriminative predictive coding networks reduces approximately to a monotone function o

arXiv:2604.21131v1 Announce Type: cross Abstract: AI-agent guardrails are memoryless: each message is judged in isolation, so an adversary who spreads a single attack across dozens of sessions slips p

arXiv:2604.21416v1 Announce Type: cross Abstract: Poisoning-based backdoor attacks pose significant threats to deep neural networks by embedding triggers in training data, causing models to misclassif

arXiv:2604.20935v1 Announce Type: cross Abstract: Wastewater treatment plants (WWTPs) need digital-twin-style decision support tools that can simulate plant response under prescribed control plans, to

arXiv:2601.22703v2 Announce Type: replace Abstract: Detecting out-of-distribution (OOD) inputs is a critical safeguard for deploying machine learning models in the real world. However, most post-hoc d

🎉 Day-0 support for @deepseek_ai V4 Pro and Flash on vLLM — a new generation of DeepSeek model, purpose-built for tasks up to 1M tokens. Alongside the release, we're publishing a first-principles walk

arXiv:2604.21428v1 Announce Type: new Abstract: Modern large-scale language model pre-training relies heavily on the single program multiple data (SPMD) paradigm, which requires tight coupling across

arXiv:2604.21006v1 Announce Type: new Abstract: We introduce Deep FinResearch Bench, a practical and comprehensive evaluation framework for deep research (DR) agents in financial investment research.

Chinese artificial intelligence developer DeepSeek today released a new series of open-source large language models. V4, as the algorithm family is called, comprises two LLMs on launch. There’s the fl

Bloomberg: DeepSeek releases its new flagship models V4 Pro and V4 Flash in preview, saying V4 Pro trails the performance of state-of-the-art models by about 3 to 6 months — DeepSeek rolled out previe

Chinese AI lab DeepSeek's last model release was V3.2 (and V3.2 Speciale) last December. They just dropped the first of their hotly anticipated V4 series in the shape of two preview models, DeepSeek-V