Find published research by company, product, platform or technology.
Showing 1–20 of 103 dossiers
GitHub Spark stops being available to existing users on August 31, 2026. Deployed apps are meant to keep running, but owners should export code to a repository now; Spark apps using `llm()` need a separate inference provider because the underlying GitHub Models service retired July 30.
The scale of the AWS–NVIDIA expansion is the headline, but the builder consequence is broader: AWS is co-engineering more of the NVIDIA stack, from CPUs and interconnects to models, vector indexing and physical-AI infrastructure, rather than merely adding another GPU instance family.
CS-4 combines three WSE-3 Turbo wafers with Cerebras’ Nexus rack design. The practical shift is architectural: compute, power and I/O become modular, while Cerebras now says the same platform is intended to support CS-5 in 2027 and a 3D-memory CS-6 generation after that.
Groq 3 LPX is moving from architecture announcement to manufactured infrastructure. Artificial Analysis measured about 3,400 output tokens/s at both 10K and 100K context on an NVIDIA-hosted private endpoint, but the single-concurrency benchmark does not yet establish public-cloud price, multi-tenant throughput or end-to-end agent speed.
A detailed engineering case study links a 68% AI-review cost reduction to prompt-cache structure and effort tuning, with security-review trade-offs still unresolved.
Google reversed its earlier ad-free approach to Personal Intelligence in AI Mode. Connected-app-derived insights may influence ads, with testing scope and advertiser controls still unclear.
DeepSeek V4.1 Flash supersedes the old Flash and Vision-Exp API lines with native multimodality, lower pricing and new architecture. Unlike those retired Flash aliases, the current DeepSeek API changelog and rate card still show V4 Pro as a distinct service.
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.
SwarmLLM does not route whole prompts to separate machines; it pipelines one model across browser tabs. A MacBook and iPhone can jointly hold Qwen 3.8 27B even when neither device can hold the full 15GB quantized model alone, with no inference server in the loop.
Meta has made the privacy-versus-price trade explicit in its Model API: developers can choose standard pricing or a contributor model ID with steeply discounted inference in exchange for training-data permission. The choice matters for proprietary code, customer data and AI SaaS workloads.
The price changes are not uniform: H100/H200 rise about 14%, B200 30%, B300 25% and GB300 about 11%. Builders using dedicated inference or training should re-run workload economics before assuming newer accelerators remain the cheapest route per completed task.
The observe–test–release loop now has explicit economics: Free and Pro include 30,000 captured generations and 25 million system-initiated AI tokens per month; Pro overages start at $1.50 per 1,000 generations and $2 per million LLM Eval/Guard tokens, while ordinary telemetry is billed separately.
beehiiv has moved AI crawler policy from a voluntary robots.txt signal to an enforceable publisher control for Max and Enterprise custom-domain sites. Its new dashboard tracks 22 AI and search crawlers, while separate structured-data and llms.txt features target AI discovery rather than access control.
Self-Hosted Machines changes the architecture of Cursor’s Cloud Agents more than another model option would. Teams can keep code, build outputs, secrets and terminal/browser actions on infrastructure they control, but the planning/inference loop remains a Cursor service and enterprise teams become responsible for worker images, scaling, secrets and production validation.
Clairic joins Clef as a specialised one-pass model on Workers AI; typed extraction has a different failure model from LLM-generated JSON, with vendor-only benchmarks and preview limitations.
The interesting part of Fastly’s AI launch is consolidation: model gateway economics, LLM security and agent-to-API authorization now sit in the same request path as the CDN/WAF infrastructure many applications already use.
The architecture matters as much as the voice quality: developers can replace a chained speech-to-text → LLM → text-to-speech loop with one full-duplex conversational model while keeping their own choice of backend reasoning model, tools and agent harness.
Meta Muse Spark Open is genuinely downloadable, unlike a promised future weight release. But 42B active parameters do not make its 816B weights cheap to host, and the Muse Model License is not Apache or MIT: model-as-a-service operators over $20M in affiliated annual gross revenue need a separate agreement.
Training experiments and batch inference can use Together AI's discounted preemptible GPUs in existing clusters. Workloads must checkpoint or requeue on interruption, and at least one standard node is required.
The post-release evidence sharpens the original story. Qwen3.8-27B can retain useful agentic-coding performance at practical 4-bit sizes, but local model quality is not a property of the checkpoint alone: quantization, reasoning effort, context handling and the agent harness can materially change the result.