#Deployment/Engineering

AISkywork's first batch of its Skywork Note AI hardware device sold out one week after launch, prompting an accelerated rollout of the wider family, including the recording clip, the Recall pendant, and the TriRing AI ring. The company says the device is meant to capture real-world conversations and moments outside the screen, so users spend less time typing and more time away from it.

AIDeepSeek says DeepSeek-V4-Flash-0731 keeps the same model architecture and size as the preview version. Today's upgrade applies only to the DeepSeek-V4-Flash API, while the DeepSeek-V4-Pro API and App/Web models remain unchanged for now. The official DeepSeek-V4-Pro release is coming soon.

AIDeepSeek has released the DeepSeek-V4-Flash API in public beta, and developers can use the latest version by setting the model name to deepseek-v4-flash. The source reports agent benchmark results far above V4-Pro-Preview, including 82.7 on Terminal Bench 2.1 and 70.3 on Toolathlon verified. V4-Flash natively supports the Responses API format and is adapted for Codex, while V4-Pro and the APP/WEB models are unchanged.

Why it matters: The release lists agent benchmark results against V4-Pro-Preview and notes Responses API support for Codex, which helps developers gauge the upgrade's practical effect on their workflows.

AIJeff Dean, Google's Chief Scientist, thanked YC partner Diana Hu for an engaging conversation at Chase Center last weekend, his first in a basketball arena. The post is a brief acknowledgment, with the surrounding context describing a Startup School 2026 discussion on AI inference hardware, the origins of TPUs, and advice for founders.

AIOpenAI cut API prices for two GPT-5.6 models, with Luna down 80% to $0.20 per million input tokens and $1.20 per million output tokens. Terra drops 20% to $2 input and $12 output per million tokens. GPT-5.6 Sol gains a Fast mode in the API that offers up to 2.5x the speed for 2x the price at the same intelligence level.

AIThinking Machines announced that its Inkling and Inkling-Small models are both available on Tinker with a limited-time discount. All Tinker models can now also be chatted with on the Tinker Playground.

AIFireworks AI ran controlled SFT experiments on Qwen3.5-9B comparing LoRA with full parameter fine-tuning across three synthetic verifiable tasks. The post argues that a FullFT advantage can come from data coverage, learning-rate tuning, or adapter rank, and it recommends testing these in that order before switching methods. Under a fixed multi-task budget, FullFT kept a 4.29-point lead over the best LoRA recipe tested, while matched data exposure favored LoRA.

AIAhmad Al-Dahle argues that AI infrastructure faces both a compute shortage and overbuilding, with the four largest hyperscalers planning roughly $725 billion of capex in 2026, up 77 percent from last year. He describes a "mutually assured construction" dynamic in which every well-capitalized player buys the same insurance against falling behind, so the industry overbuilds by construction.

AIBerkeley AI Research extended the K-Search evolutionary kernel framework with an MLX backend and a CUDA-to-MLX translation layer, letting it adapt existing CUDA kernels for Apple Silicon. The team reports a 0.97x speedup relative to the native MLX Attention kernel and up to a 20x prefill speedup over the community mlx-lm implementation on the Mamba SSM kernel. The method uses Gemini 3.5 Pro Preview to both reason about optimizations and write candidate kernels.

AIAugment Code has made GPT-5.6 Sol the default model in Cosmos, choosing it as the most token-efficient model to clear its pass-rate floor for long-horizon software engineering tasks. The company ranks models by cost per task rather than list price per million tokens, since retries on failed steps add token spend. Users can still select any model, and the default will change as more token-efficient models emerge.

AIFireworks AI describes fine-tuning Qwen3-Embedding-8B on private (query, positive) pairs using bidirectional InfoNCE loss through its Training SDK, then serving the model via an OpenAI-compatible embeddings endpoint. The post reports that around 150 training steps was enough, that rank-32 LoRA landed within about one point of full-parameter fine-tuning, and that gains were largest where the base model struggled, while tasks like CoSQA and FiQA2018 showed flat results.

AILTM has partnered with Cognition to deploy Devin, the AI software engineer, through BlueVerse RightLogic, a managed, outcome-based service that clears customers' vulnerability backlogs. RightLogic is designed to clear 80 percent of an enterprise's CVE backlog, up from the 60 percent previously delivered, and will focus first on banking, financial services, and insurance. The service is the first of five joint offerings the companies plan to bring to market.

AIJetBrains tested the ponytail skill for Claude Code across 80 paired tasks and found a median 10.3% cost reduction, with p=0.004. Code written fell about 15% median versus the advertised 54%, reaching 31% on larger builds and little on already-lean tasks. No quality difference was detected, and the skill only self-activated when its ruleset was injected by a plugin hook.

Why it matters: The benchmark separates advertised savings from measured results and shows the code cut depends on how much the baseline agent over-builds.

AICursor says its India user base tripled in the past year and that Indian users make more agent requests per user than any other country. It is launching Cursor Start, a ₹649/month plan with generous access to Grok 4.5 and Composer for Indian developers.

AICyera is joining with Oasis Security, which builds agentic access management for non-human identities such as API keys, service accounts, OAuth tokens, and agent credentials. The combination pairs Cyera's knowledge of where sensitive data lives with Oasis's visibility into which identities and agents can reach it. Sequoia Capital, which backed both companies since their Series A rounds, says the pairing covers the full path an AI agent takes through an enterprise.

AILiquid AI released LFM2.5-Encoder-230M and LFM2.5-Encoder-350M, bidirectional encoders built on the LFM2 hybrid architecture and available on Hugging Face. They support an 8,192-token context and are designed for fine-tuning on classification and token-level tasks. On CPU, LFM2.5-Encoder-230M is the fastest model tested from 1K tokens up, running about 3.7x faster than ModernBERT-base at 8,192 tokens.

AIMoonshot AI has open-sourced MoonEP, a high-performance communication library for distributed MoE workloads. The library is designed to reduce communication overhead in expert-parallel training and inference at scale. The code is available on GitHub under MoonshotAI/MoonEP.

AIMoonshot AI's Kimi, in collaboration with kvcache-ai, has open-sourced AgentENV, a distributed system for running agent environments at scale. Its components power agentic RL training for Kimi K3, supporting fast snapshot, resume, and fork for large-scale parallel agent workflows. The project is available on GitHub at

AIMoonshot AI has open-sourced FlashKDA, a high-performance CUTLASS-based implementation of Kimi Delta Attention kernels. It delivers a 1.72×–2.22× prefill speedup over the flash-linear-attention baseline on H20 and works as a drop-in backend for flash-linear-attention.

AISafe Superintelligence (SSI) announced a long-term strategic partnership with NVIDIA, which is making a substantial investment in the company. The investment is expected to let SSI increase its compute tenfold within the next 12 months.

AIThe University of Pittsburgh's RAMMP team is integrating Meta's DINOv3 and SAM models into on-device assistive robotics to detect door buttons, cups, and curbs for navigation assistance. The models run on compact, battery-powered hardware, with optimizations such as reduced memory footprint and lower precision, enabling real-time perception without network connectivity. RAMMP's perception system pairs SAM-based auto-labeling with an RF-DETR detector fine-tuned on DINOv2 embeddings, and the team is now testing voice and touch input for object selection.

AIFireworks AI made the open-weight Kimi K3 available for inference and training on its platform, with US-only serverless endpoints and Zero Data Retention. In its own head-to-head with Opus 5, the post reports K3 at 92.7% accuracy and $0.52 per task on SWE (480) against Opus 5's 94.8% and $1.05, with the vendor claiming up to 5x better cost efficiency per task.

Why it matters: The post compares Kimi K3 with Opus 5 on accuracy and cost per task, giving readers concrete figures to judge the open model against closed alternatives for their own workloads.

AIFireworks AI has made Kimi K3 available for Multi-LoRA serving and training in private preview through Fireworks Serverless Training. The post explains how small LoRA adapters can be trained on K3 and served with live merge or multi-LoRA deployment, and it reports two example tasks, Countdown and Frozen Lake, with reward curves.

AITrilogy released an AI Cybersecurity Playbook built around Kimi K3, served on Fireworks through an OpenAI-compatible API. The playbook pairs deterministic tools for indexing, parsing, and deduplication with Kimi K3 for reachability, impact, and remediation judgment, followed by a separate verification stage.

AILangChain Blog argues that companies need to own their AI intelligence rather than rely on generic models, because general models do not know company-specific policies, workflows, or risk tolerances. Ownership means controlling the agent system (model optionality, harness, and context), the economics, quality, and risk of AI work, and how intelligence compounds over time. The post uses an insurer's claims processing as an example of why off-the-shelf models fall short.

AIBerkeley's STAR Lab packaged multiple LLM-based "autoresearch" algorithms into a single API within the GEPA package, letting users mix and match them. The optimizers can be applied to tasks including prompt writing, agent design, and code optimization. The quoted thread adds that GEPA, AutoResearch, and Meta-Harness each win on different tasks, and that the new optimize_anything omni meta-optimizer beats every standalone optimizer at a matched budget.

AITri Dao congratulated the Etched team, calling it strong in both hardware design and kernels. The quoted Etched post says the company raised $300M in Series C funding at a $10.3B valuation to accelerate production of its inference clusters. It also opened an 80,000-sqft, 10-MW facility for production and prototyping.