Radar / Categories / AI infrastructure
New AI infrastructure, ranked by traction
New routers, memory layers, tokenizers and inference infrastructure. Every launch from the last few weeks, ranked by JEV traction and GitHub stars. Updated four times a day.
Router · 18
Modular open-sources MAX inference server, Mojo stdlib and accelerator kernelsThe core of Modular's unified AI deployment stack is now open: the MAX inference server (OpenAI-compatible endpoints)…infra · JEV 0.76OmniRouteFree MIT AI gateway: one endpoint, 290+ providers and 500+ models with quota-aware auto-fallback.infra · JEV 0.75DeepSeek open-sources Ascend infrastructure stackDeepSeek published Ascend-optimized versions of its NVIDIA-proven infra components — TileLang Ascend, DeepGEMM-Ascend…infra · JEV 0.65DeepSeek Elastic Compute (DSec): Sandbox Infrastructure for Agentic Training at ScaleDeepSeek's production sandbox platform behind all its RL training from V3.2 through V4.1: a unified SDK over FnCall…infra · JEV 0.6Perplexity PhotonRust-based retrieval engine powering Perplexity's new Fast Search API, cutting p99 latency from 800ms to 65ms.infra · JEV 0.54Gimlet Labs x Cerebras: ultrafast inference cloud at 3,000 tokens/secGimlet Labs and Cerebras (NASDAQ: CBRS) partner on a disaggregated inference cloud combining wafer-scale compute with…infra · JEV 0.48Magnitude: self-optimizing open-source inference engine for local agentsOpen-source inference engine (YC S25) that profiles your machine, recommends the best open models for it, and tunes…infra · JEV 0.38Arm AI Portal: pre-optimized models for Arm hardware, with MCP agent accessDedicated AI portal giving developers a library of pre-optimized models (language, voice, video) benchmarked for Arm…infra · JEV 0.38OpenAI Pro 500: $500/month tier with Ultrafast compute for GPT-6 AstraOpenAI announced a $500/month Pro 500 tier at DevDay (Sep 29) with 'Ultrafast' compute for GPT-6 Astra in ChatGPT Work…infra · JEV 0.36nestrilabs/virtio-nvgpuExperimental virtio device giving KVM virtual machines near-native NVIDIA GPU access; a March 2026 repo that hit the…infra · JEV 0.28Liner Model API: Cost-Aware Model RoutingAPI that automatically routes each LLM request to the cheapest model capable of handling it; Liner claims over 50%…infra · JEV 0.28SuperX AI Token Platform: unified multi-model access metered in tokensSuperX (NASDAQ: SUPX) launched an AI Token Platform giving unified third-party model access through one interface with…infra · JEV 0.25CometAPI — unified OpenAI-compatible API for 500+ AI modelsA single API key and OpenAI-compatible interface to access 500+ frontier and multimodal AI models; launched Oct 1, 2026.infra · JEV 0.24MCP Nexus 1.0: routing and discovery layer for MCP toolsOne umbrella MCP endpoint that sits in front of hundreds of tools and exposes only the few capabilities an agent needs…infra · JEV 0.22LiteLLM Lens — AI agents that analyze agent traces inside the gatewayTrace-analysis product using AI agents to find recurring failures across agent runs: send OpenTelemetry traces to the…infra · JEV 0.22OpenClaw Enterprise: free, MIT-licensed control plane for persistent agentsThe open-source agentic framework goes enterprise: vendor-neutral control plane with multi-tenancy, hard security…infraAG-UI 1.0Stable agent-to-application event protocol with schema-generated TypeScript, Python and .NET SDKs.infralaya: calibrated classification layer for routing and guardrailsCompact text-classification model for routing, scoring, guardrails and moderation, trained for calibrated decisions.infra
Memory · 12
Cloudflare Agents Week: Sandboxes GA, 50K concurrent Workflows, Managed OAuth for agentsA dozen agent-infrastructure launches in one week: persistent Linux Sandboxes (GA) with real shell/filesystem/state…infra · JEV 0.69ai-memory: long-term memory for agent coding CLIsRust solution for long-term memory for agent coding CLIs, facilitating handoff between different agents and sessions.infra · JEV 0.68Hindsight: agent memory that learnsAgent memory system built for learning over time: retain/recall/reflect operations with SOTA scores on the LongMemEval…infra · JEV 0.63Alibaba Zhenwu V900 AI acceleratorAlibaba's T-Head chip unit unveiled the Zhenwu V900 AI chip at the Apsara Conference, claiming 3x the performance of…infra · JEV 0.59Context Mode: context-window optimization for AI coding agentsMCP server for AI coding agents that sandboxes verbose tool output and persists session memory, cutting context use by…infra · JEV 0.57OpenObserve v1.0: GA with AI Observability for AgentsOpen-source observability platform hits 1.0, headlined by AI Observability: agent tracing, LLM monitoring, evals, and…infra · JEV 0.54dlab Open Source WeekTim Dettmers' lab released an open ecosystem for frontier AI on own hardware: an inference framework running Qwen 3.8…infra · JEV 0.54Docker Sandbox Kit: coding-agent sandboxes donated to the CNCFDocker is taking its Sandbox Kit spec to the CNCF - a standard for packaging coding-agent sandboxes (and their…infra · JEV 0.51MOI-1A orbital computing satelliteTakeMe2Space launched India's first orbital computing satellite — customers upload containerized AI models and it…infra · JEV 0.48tpu-megakernels (Inferact): open fused kernels that beat GB200 on Kimi K3Open-source fused 'megakernels' for Google TPU v7 that serve Moonshot's Kimi K3 at 709 tokens/sec (and Qwen3.8-27B at…infra · JEV 0.46okf-agent-memoryGit-native persistent memory for AI coding agents implementing Google OKF v0.2.infra · JEV 0.14PageIndex: vectorless, reasoning-based RAGMIT-licensed RAG engine that replaces the vector DB with a hierarchical tree index and lets an LLM reason its way…infra
Tokenizer · 1
Other · 1
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.