Orchestration · API

Application framework for orchestrating LLM workflows, tool calling, retrieval, and agents across multiple providers in Python and TypeScript ecosystems.

Orchestration · API

LangGraph is a library for building stateful, cyclic agent and workflow graphs on top of LangChain—suited to multi-step tools, human-in-the-loop approvals, and durable execution patterns that go beyond linear chains.

LLM observability · API

LangSmith is LangChain's framework-agnostic platform for tracing, debugging, evaluating, prompt testing, and deployment workflows for AI agents and LLM applications.

Data framework · API

Data framework for LLM applications focused on ingestion pipelines, indexing, retrieval, and query orchestration over private and enterprise content sources.

AI platform · API

Microsoft Foundry is Azure's platform for discovering models, building agents and generative AI applications, evaluating them, and applying Microsoft identity, networking, governance, and observability. Microsoft documentation describes a catalog of more than 1,900 models; exact model and regional availability varies.

Image generation

Midjourney is an AI research lab and creative platform known for image and video models used to generate highly stylized visual content.

data · API

An open-source vector database designed for high-performance similarity search and analysis of large-scale vector data. It handles millions of vectors efficiently with a query latency of under 100ms for similarity searches.

Compute · API

Serverless compute platform for AI inference and batch workloads, offering GPU execution, scalable workers, and code-first deployment patterns for model-powered applications.

AI agent

Muse is Meta's personal agent. Meta's help center says you can ask it questions, set reminders, continue a conversation, and connect it to apps so it can take actions. Connect 2026 says the current surfaces are the Muse app, WhatsApp, the web, and the Mac desktop app. It is free with a usage limit, and Meta offers a paid subscription when that limit is reached. Connectors can read services such as calendar or email. By default Muse asks before actions such as sending an email.

Developer tool

Muse Code is Meta's coding agent for the terminal and CI, built on Muse Spark. The official installer places a muse binary on macOS and Linux. Interactive sessions plan, edit, and run commands with workspace trust, approvals, and an OS sandbox. Headless muse exec runs one prompt for scripts and CI. Muse Spark 1.3 is available; Meta's Muse Code overview still documents muse-spark-1.2 as the default until you change it. Billing is Meta Model API pay-as-you-go or a Muse Code monthly subscription.

Workflow automation · API

n8n is a workflow automation platform for technical teams, with visual workflows, code-level flexibility, integrations, self-hosting options, and AI agent/RAG use cases.

productivity · API

Local model runtime for running and serving open LLMs on developer machines and private infrastructure, with simple pull/run workflows and API access.

Developer tool · API

The OpenAI Agents API is the September 10, 2026 public-beta API that exposes the Codex harness as a managed cloud runtime. OpenAI runs sessions, orchestration, context compaction, and recovery. Your application supplies tools and chooses the environment: an OpenAI-hosted sandbox or a self-hosted executor. Agents can run commands, edit files, call MCP servers, spawn subagents, and resume durable sessions. Requests use the OpenAI-Beta: agents=v1 header. This is not the Codex CLI product and not the in-process Agents SDK.

Developer tool · API

Codex is OpenAI's coding agent for the ChatGPT desktop app, ChatGPT Work on the web, the Codex CLI, and the IDE extension. The official models page recommends GPT-6.1 Sol for repeated long-running work when the account has it, Astra for the hardest end-to-end work, and GPT-6 Luna for focused high-volume tasks. Codex Cloud runs tasks in a published environment with its own workspace, including while the computer is asleep. CLI example commands on the models page use gpt-6.1-sol, gpt-6-astra, and gpt-6-luna.

IDE · API

Provider of widely used frontier model APIs for text, vision, and audio, with strong developer tooling and broad ecosystem adoption across production AI applications.

Model gateway · API

OpenRouter aggregates access to many foundation models behind one API and billing surface, letting teams route prompts across providers for cost, capability, or failover without maintaining separate SDKs and accounts for every vendor.

Vector database · API

Managed vector database for semantic search and RAG systems with metadata filtering, namespaces, and cloud-hosted reliability for production retrieval workloads.

Evaluation · API

Promptfoo provides AI security testing, red teaming, guardrails, model security checks, and evaluations for prompts, models, agents, and RAG pipelines.

Vector database · API

Vector database focused on high-performance similarity search with strong payload filtering, hybrid retrieval features, and both open-source and managed cloud options.

Developer tool · API

QoderWake is Alibaba Qoder's always-on digital-employee runtime. Official 1.0 notes (September 3, 2026) describe Wakers — named roles with memory, skills, connectors, and permissions — that take work from chat, IM @mentions, schedules, hooks, or API calls. Teams orchestrate multi-role work with Groups and WakerFlow. The product page lists Mac, Windows, and cloud deployment; docs cover a local Web Console. This is not Qwen Code and not the older Qoder IDE.

Developer tool · API

Qwen Code is an open-source terminal coding agent from the Qwen team with documented file operations, shell commands, web access, MCP integrations, and subagents. It can use Qwen models and compatible providers, making it useful for teams that want a modifiable coding-agent surface.

Vector database · API

Redis Vector Search extends Redis with vector similarity queries alongside familiar key, JSON, and search capabilities—useful when you already run Redis for caching or features and want co-located embeddings with low-latency hybrid retrieval without adding a separate database cluster.

Model hosting · API

Managed inference platform for running open and custom models through simple APIs, with usage-based billing and strong support for image, video, and multimodal workloads.

AI infrastructure · API

Router.com is Ramp's multi-model API for accessing, evaluating, and routing AI workloads without rebuilding an application for each provider. Its launch includes flex-tier routing, shadow-model evaluation, benchmark-based selection, NVIDIA Switchyard escalation, and per-request cost and latency visibility.