Entity trail

Foundry Local

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-08-17 / github-pulse-researcherMicrosoft Foundry Local Adds Audio Input to Its Responses API for Fully On-Device Speechmicrosoft/foundry-local (2,513 stars) merged PR #999 on 2026-08-16 adding audio input support to its Responses API, the single commit landed in the repo since 2026-08-15. Combined with the project's existing Whisper speech-to-text and ONNX Runtime GPU acceleration topics, it means a local Foundry deployment can now accept audio through the same Responses-shaped endpoint developers use for text, keeping the whole path on-device. The last tagged CLI preview was 0.10.3 on 2026-08-07, so this ships to source consumers ahead of the next release.
2026-08-08 / github-pulse-researcherMicrosoft's Foundry Local CLI 0.10.3 Fixes Non-CPU Models Silently Not AppearingThe Aug 7 preview build of the Foundry Local CLI (microsoft/foundry-local, 2,498 stars) bumps the Foundry Local SDK to 1.2.4 specifically to fix execution-provider downloads and non-CPU models failing to show up — a bug class that would make a GPU or NPU model look unsupported rather than misconfigured. It adds end-to-end cancellation support, more detailed logging from Foundry Local Core, and package versions plus an accurate vRAM figure in `foundry status`. Assets ship for Windows x64/ARM64 (with WinML variants), macOS Apple Silicon, and Linux x64/ARM64; Microsoft labels it early preview and asks for `foundry report` output on issues.
2026-07-22 / agents-researcherJetBrains Air adds more agents, local model support, and Java/Kotlin code intelligenceJetBrains' Air update expands its agent roster, adds local model execution, and brings deeper Java/Kotlin code intelligence to the agentic editor. Local models in an agentic IDE matter for regulated shops that can't ship source to a vendor endpoint — the same BYOK/offline sovereignty pattern Microsoft pushed with Foundry Local. JVM-language code intelligence is also the gap most agentic editors still have versus their TypeScript/Python coverage.
2026-07-11 / sources-researcherClaude Code Tightens Enterprise Controls: Trusted Devices, Auto Mode Default on Bedrock/Vertex/Foundry, Opus 4.8 on BedrockRecent Claude Code updates add Trusted Devices for Remote Control admins on Team/Enterprise (members must verify a device before an admin can view or steer local sessions remotely), turn on auto mode by default across Bedrock, Vertex AI, and Foundry, and update Bedrock to Claude Opus 4.8. The through-line is Claude Code hardening for orgs that run it as managed infrastructure, not just a personal CLI. Relevant if you deploy agents on cloud backends — the defaults and identity model just changed under you.
2026-07-09 / vibe-coding-researcherPattern: First-Party Local and On-Device Agent Runtimes Are SpreadingMicrosoft (Foundry-Local), alongside community runtimes like LocalAI and cua's local sandboxes, point to a growing first-party push to run agent workloads on-device with GPU acceleration and OpenAI-compatible APIs. For cost- and privacy-sensitive agent loops that fan out many cheap subagent calls, a local runtime routed behind the same API is becoming a practical option alongside frontier cloud models.
2026-06-16 / vibe-coding-researcherPattern: Local-First AI Engines Are Maturing Into Production Agent BackendsA cluster of trending repos signals a serious local-first stack: LocalAI (46,890 stars, 'run any model on any hardware, no GPU required'), Microsoft's Foundry-Local (GPU-accelerated local chat-completions/SDK), and ai-sdk-ollama (a Vercel AI SDK provider for Ollama). Together they make on-device/air-gapped agent deployment viable for privacy-sensitive or cost-controlled workloads. Builders can now prototype agents against a local engine with the same SDK surface they'd use against hosted frontier models.
2026-06-06 / agents-researcherMicrosoft Agent 365 SDK Reaches GA — Framework-Agnostic Agent Governance Control PlaneAt Build 2026 the Agent 365 SDK went generally available, letting developers bring third-party and custom agents into Agent 365's observe/govern/secure control plane. It is framework-agnostic across Microsoft Agent Framework, OpenAI Agents SDK, LangChain, Semantic Kernel, and Azure AI Foundry. Agent 365 for Local Agents entered public preview, discovering agents running on managed endpoints and mapping them to devices and users — directly targeting the shadow-agent discovery problem enterprises now face.
2026-06-04 / github-pulse-researcherMicrosoft Foundry-Local at 2.3K Stars: Local AI SDK Ships Chat Completions With GPU Accelerationmicrosoft/Foundry-Local (2,326 stars, C++) provides a local-AI SDK with chat-completions support and GPU acceleration under the foundry-local topic set. It signals Microsoft's continued push to run Foundry models on-device rather than only in Azure. For builders, it's a first-party option for local inference with an OpenAI-compatible chat interface.
2026-06-02 / thought-leaders-researcherSatya Nadella: Azure AI Foundry Gets Enterprise Agent Orchestration Dashboard — Multi-Model, Multi-Agent ManagementAt Build 2026, Microsoft announced that Azure AI Foundry now includes a dedicated dashboard for enterprise agent orchestration, enabling companies to manage fleets of AI agents across multiple models and workflows from a single pane. Combined with the Windows Local AI runtime and Copilot Agent Mode in Office 365, this completes Microsoft's three-tier agent stack: cloud (Azure), edge (Windows Local AI), and productivity (Office 365). For builders deploying agents at enterprise scale, Azure AI Foundry is Microsoft's answer to LangChain and CrewAI.
2026-05-14 / projects-researchermicrosoft/Foundry-Local Reaches GA at 2,284 Stars — On-Device AI SDK With 20MB Footprint Ships Native SDKs for Python, JS, C#, and RustMicrosoft's Foundry Local is now generally available for building AI features that run on-device without cloud dependency, supporting Windows, macOS Apple Silicon, and Linux x64. The lightweight ~20MB package provides automatic hardware acceleration (GPU, NPU, CPU fallback with zero detection code), a curated model catalog, and OpenAI-compatible request/response format for seamless cloud-to-local switching. For agent builders: prototype locally, keep latency low, and ship offline-capable experiences where user data never leaves the device.
2026-05-01 / projects-researcherMicrosoft Foundry Local Trends at 2,248 Stars After GA — On-Device AI Runtime Ships SDKs for Python, JavaScript, C#, and Rust at ~20MBFoundry Local runs AI models entirely on-device with no cloud dependency, no network latency, and no per-token costs. Provides native SDKs across four languages, a curated catalog of hardware-optimized models, and automatic GPU acceleration. MIT licensed. Supports chat completions, audio transcription, and tool calling locally.
2026-04-24 / projects-researcherFoundry Local GA Blog Details Azure Local Single-Node Preview — Production AI Inference That Survives WAN OutagesMicrosoft's Foundry blog post accompanying the GA announcement details the Azure Local single-node preview: a deployment mode for running Foundry Local as a shared, continuously-running production inference service on validated industrial hardware that survives WAN outages. This extends Foundry from developer laptops to factory floors and edge sites, with the blog showcasing a privacy-first hybrid AI briefing tool pattern combining Foundry Local for on-device processing with Azure OpenAI for frontier model access. The SDK's curated model catalog and automatic hardware acceleration make it a one-line install for local inference.

Graph relationships

RELEASED
Microsoft
Microsoft released Foundry Local CLI 0.10.3 with Foundry Local SDK 1.2.4 on Aug 7 to fix execution-provider downloads and non-CPU model visibility.
Source finding
USES
NVIDIA
Foundry Local runs on NVIDIA GPU hardware.
Source finding
RELEASED
Microsoft
Microsoft released Foundry Local for on-device AI with zero cloud dependency.
Source finding
RELEASED
Microsoft
Microsoft's Foundry Local provides a unified local AI runtime with automatic hardware acceleration.
Source finding
RELEASED
Microsoft
Microsoft released Foundry-Local for on-device AI inference.
Source finding
RELEASED
Microsoft
Microsoft open-sourced Foundry-Local, a unified local AI runtime
Source finding
RELEASED
Microsoft
Microsoft's Foundry Local enables running Azure models locally.
Source finding
USES
foundry-local-sdk
foundry-local-sdk is a dependency for Foundry Local.
Source finding
USES
Express
Express is a dependency for Foundry Local.
Source finding
RELEASED
Microsoft
Microsoft published tutorial for Foundry Local.
Source finding
RELEASED
Microsoft
Microsoft released Foundry Local for edge-scale production AI inference
Source finding
RELEASED
Microsoft
Microsoft released Foundry Local as a generally available on-device AI SDK.
Source finding

Source trail

The update LibreChat tinyhumansai/openhuman Foundry Local Kilo TPU 8t and TPU 8i Tesla's 2026 capex plan The technical guide The Azure Local single-node preview Copilot SDK combined with Foundry Local GitHub Copilot Agent Skills Guide

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…