Find published research by company, product, platform or technology.
Showing 1–20 of 48 dossiers
Production integrations using GPT-4o, GPT-4o mini, GPT-4.1 or GPT-4.1 mini dated snapshots need model inventory, tool-call regression testing and migration before January 14, 2027.
Kimi K3.1 Pro ships via API at $2 input/$10 output per million tokens, with a 2–3x faster Highspeed tier. Independent Artificial Analysis testing scores 55 on its Intelligence Index and 67% on Terminal-Bench 4.0. Open weights remain promised, not released.
The useful part of Smaug Agentic is not another frontier-style benchmark claim. Abacus.AI is publishing a drop-in Kimi K3 derivative that targets a specific production failure mode in coding agents: long runs that burn the reasoning budget without converging. The weights and model card are public, but the training data is not disclosed and the benchmark gains remain vendor-produced.
The material issue is not ordinary model distillation. Anthropic’s evidence suggests a customer-facing AI product may have used a rival model as an undisclosed backend while simultaneously harvesting those interactions for training, turning routing architecture into a privacy and trust boundary.
GitHub Apps no longer need repository-administration permission just to create repositories. New fine-grained installation controls and token-rotation support reduce the privilege required for integration automation.
Bounded decision models are turning into a real model category. Cloudflare's entry is open-weight, multimodal and Jev-API compatible, while its fastest variant is aimed at latency-sensitive agent routing.
Developers can now train and evaluate adapters through custom code while Together runs the GPU-heavy operations. The flexibility comes with B200-only support, a $100 credit threshold and ongoing charges for idle ready resources.
This is a hard capability removal rather than a routine model migration. Products built on OpenAI’s video-generation API now need another provider or a redesigned video path because the official deprecation table offers no successor endpoint.
CLM-8B targets the same narrow decision layer as Jev, but with open weights, local deployment and a contrastive architecture that separates state and action representations. The headline speed and coding results are researcher-produced and need careful interpretation.
This is not a normal ranking update. Google is changing the structure of commercial search results in the EEA under the Digital Markets Act, creating explicit result surfaces for vertical search services and suppliers that do not appear the same way elsewhere.
The two August 28 changes move a common production-agent problem out of bespoke application code: builders can derive memory boundaries from authenticated JWT claims, enforce them with Cedar policy, and organize the stored memory using runtime tenant dimensions.
The Assistants API shutdown date has passed. OpenAI’s deprecation documentation lists August 26, 2026 as the removal date and directs developers to Responses and Conversations for replacement workloads.
Sentence Transformers 6 now has both unified multi-vector inference and a documented end-to-end training workflow. A new project-authored benchmark shows fast domain adaptation on a single GPU, but the result is workload-specific and index costs remain high.
Meta’s Muse Glimmer 30B combines tool use, coding, vision and agentic task completion with official local-runtime artifacts. A 17GB GGUF build targets 24GB-VRAM machines, but Meta also attaches a separate usage policy, so builders should distinguish weight availability from unrestricted use.
Jalapeño is working first-party silicon rather than a roadmap item, and OpenAI now says AI itself materially accelerated the design process. The distinction still matters: tape-out means the design was finalized for manufacturing; it does not mean fleet-scale production qualification or API deployment is complete.
Hy4 preview is a very large sparse model with public full and FP8 weights, native speculative decoding and a 1M-token context path. Its open release makes Tencent’s claims testable, while the 1.56TB full checkpoint keeps self-hosting firmly in server-scale territory.
Jev made bounded decision models visible; Strands Decider makes the pattern reproducible inside an agent stack. AWS replaced Qwen3.5-2B's language-generation head with a small scoring head and released the recipe, creating a local alternative for decisions that do not need a full generative model.
K2 Horizon is notable less for another benchmark claim than for reproducibility: IFM is publishing model weights, architecture, training code, data or construction recipes, evaluation resources and intermediate training material instead of stopping at a final checkpoint.
MiMo-V2.6 is more useful than another benchmark launch because builders get both capable multimodal weights and a rare view into the reinforcement-learning machinery that produced them: code, environments, run costs and even failure notes from the training cluster.
GitHub’s credential-response story now has both discovery and containment: enterprise owners can export SSH keys, PATs, OAuth and GitHub App tokens with ownership, scope and last-use metadata, then use selective revocation rather than invalidating every credential a user holds.