Find published research by company, product, platform or technology.
Showing 1–20 of 205 dossiers
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.
DeepSeek V4.1 Flash supersedes V4 Flash and Vision-Exp on the hosted API, keeps native multimodality, reduces serving costs through a smaller active path and KV cache, and introduces a transition in which V4 Pro traffic will temporarily route to V4.1 Flash at V4.1 Flash rates.
Fusion is interesting less as another routing feature than as a different agent-cost architecture: two persistent model contexts divide planning, review and execution instead of making one expensive model handle every token. The practical question for builders is shifting from token price to cost per completed task.
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.
OpenAI’s internal data turns “agents make researchers faster” into a measurable operating model: heavy concurrent agent use, record experiment throughput and rising task complexity, alongside high token spend and persistent human intervention on longer work.
The important change is economic rather than another flagship benchmark win. OpenAI is making capable agent and coding workloads materially cheaper, with Luna approaching older Sol-class results at a tiny fraction of the task cost and GPT-6 prompt caching discounting reused input by up to 90%.
The interesting change is economic as much as benchmark-driven. Anthropic is compressing capability that previously justified its larger Fable tier into Opus pricing, while cutting Opus list prices and expanding immediate availability across the major clouds.
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.
The material change is that model routing is no longer a single opaque optimization target. Developers can now tell Copilot whether to bias Auto toward lower cost, a middle ground or higher quality while GitHub still chooses a model prompt by prompt.
Cloudflare Workflows now prices steps and persisted state on paid plans, making workflow structure and retention part of the cost calculation for durable jobs and AI automation.
Notion Workers are now metered inside the same credits system as Custom Agents. The important builder shift is that schedules, webhook fan-out and agent tool-call counts now directly affect cost.
The Agent Host is becoming less like a background chat process and more like durable development infrastructure.
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.
The important shift is that agent orchestration itself becomes a managed API surface: context compaction, tool discovery, programmatic tool calls and subagent coordination can now come from OpenAI’s maintained Codex harness rather than an application team rebuilding those layers.
Agent Identity is moving from a standalone credential boundary into a mainstream serverless runtime. Cloud Run can now assign agent identities and register agents/MCP servers automatically, reducing custom discovery and identity plumbing while keeping the runtime integration itself in Preview.
Railway Cloud Agents are managed, persistent development machines rather than a new model or harness. They reuse developers’ existing agent credentials, sleep when disconnected by default, retain disk state, and live inside Railway project environments—blurring the boundary between remote coding workspace and deployment platform.
Agent Plugins 1.0 now has documented support across VS Code, Cursor, GitHub Copilot, ChatGPT/Codex, Kiro and several open-source agents. That makes the format materially more useful for cross-client distribution, but portable components remain limited to Agent Skills and MCP servers while permissions, hooks, commands and host UX stay client-specific.