A self-custodial wallet for AI agents with immutable spending policies, transaction protection, gas abstraction, and support for autonomous onchain activity.

Alibaba's open-weight 27B dense companion to Qwen3.8-Max: Apache 2.0 license, multimodal image-text input, built for local deployment and small-batch inference.

Self-hosted AI gateway that sits in WhatsApp, Telegram, Discord, Slack, and more. MIT, ~176k GitHub stars, npm `openclaw`.

Meta's open-weight 30B agentic multimodal model distilled from Muse Spark, designed to run always-on coding and tool-use agents locally on a single consumer GPU.

MemGPT's company and OSS agent memory runtime. GitHub letta-ai/letta is ~20.5k stars. Cloud is $20 and $200 per project plus usage.

Open-source browser API for agents. Cloud Launch is $0 plus usage with a one-time $30 credit. Scale is $250/month plus usage.

Open-source coding agent for VS Code and JetBrains. Apache-2.0, local models, CLI, and MCP. Acquired by Cursor; the repo stays public.

Self-hosted coding assistant. Rechecked: Community $0 (up to 5 users); Team $19/seat/month (up to 50). Apache-2.0 core, not a free cloud Copilot.

Open-source self-hosted AI personal assistant for managing email, calendar, tasks, and workflows. Privacy-focused, cross-platform support. Created by Peter Steinberger, formerly known as Clawdbot.

Local models plus Ollama cloud. Rechecked: Free $0, Pro $20/month or $200/year, Max $100 paused, Team $25/seat (5-seat min). Not a 30-minute Llama3 tutorial.

Postgres vector extension. Rechecked: PostgreSQL license, 22,651 stars. Not a Faiss wrapper. No hosted dollar seat.

Cloud-native ANN engine. Rechecked: Apache-2.0, 1,716 stars, vald.vdaas.org. Not an unaudited fastest-NGT census.

OSS search+vector engine. Rechecked: Apache-2.0, 7,052 stars. Cloud has a free trial; unit table exists. Not an unpriced toolkit.

Open relational DB. Rechecked: PostgreSQL License, 21,814 stars, 18.6 current. Not a leftover BSD slogan page.

Oracle relational DB. Rechecked: Community GPLv2, 12,382 stars, 26.7.0 tree. Not a leftover social-media origin story.

Cross-platform team chat where AI agents are first-class teammates. Cloud pods or BYOA Claude Code/Codex. Invite-only preview is free.

Self-improving open-source AI agent by Nous Research with a built-in learning loop, multi-channel gateway, and model-agnostic runtime. MIT licensed.

Token-efficient microkernel AI agent with ML-based smart routing, persistent memory, a layered sandbox, built-in search, and local embeddings. Apache 2.0.

Free MIT AI router and token saver that connects Claude Code, Codex, Cursor, and Cline to 40+ providers with auto-fallback and RTK token compression.

MIT local gateway that points Claude Code, Codex, Pi, OpenCode, or Cline at 48 providers. Rechecked 45,953 GitHub stars on 2026-08-19.

AI-native GPU cloud. Rechecked: NVIDIA Reference Platform Partner, H100 from $2.00/GPU-hr, H200 $2.60, B200 $4.00. No invented seat.

Open-source agent-native backend: Postgres, auth, storage, edge functions, compute, hosting, and a model gateway driven by CLI and MCP.

Free MIT AI gateway with one endpoint, 339 providers, 90+ free tiers, quota-aware fallback, and token compression for Claude Code, Codex, and Cursor.