Recently Added AI Tools

DeepSeek's flagship 1.6T MoE model with 49B active parameters, 1M-token context, MIT open weights, and world-leading coding scores at a fraction of closed-model prices.

SpaceXAI's frontier 1.5T reasoning model with 500K-token context, agentic coding strength, and real-time X data at roughly half the price of rivals.

Zhipu AI's newest flagship built for coding and cyber defense, squeezing 50% more coding capability out of the GLM-5 base through post-training scaling, ranked #1 among open-source models on Terminal Bench 3.0 and Agents' Last Exam (CLI).

SpaceXAI's always-on AI teammates that run on their own cloud computer, sign into your tools, and finish work end-to-end — available in beta for SuperGrok Heavy and Cursor Ultra subscribers.

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.

MiniMax's open-weight music generation model with studio-grade sound, natural vocal synthesis, and full-song generation from prompts and lyrics, with official ComfyUI support.

Free open-source desktop app to run and train AI models 100% locally, with Claude Code and Codex agent integration, media generation, and private web search.

Spotify's vendor-neutral agentic development environment. One place to launch, manage, and switch between Claude Code, Gemini CLI, and Codex sessions, with per-session Git worktrees and Portal-powered organizational context.

Programmable account and virtual wallets for the agentic Internet, combining stablecoin micropayments, delegated identity, and policy-based spending controls.

Open-source agent runtime from Cloudflare that gives every AI agent a durable SQLite-backed filesystem and switchable execution environments (container, isolate shell, or JavaScript isolate) running entirely on Workers.

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

DeepSeek's flagship 1.6T MoE model with 49B active parameters, 1M-token context, MIT open weights, and world-leading coding scores at a fraction of closed-model prices.

DeepSeek's official small agent model with 284B total / 13B active MoE parameters, 1M-token context, MIT open weights, DSpark speculative decoding, and $0.14/$0.28 per 1M tokens pricing.

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.

Alibaba's flagship 2.4T MoE model with 95B active parameters, 1M-token context, native multimodal input, and top-ranked agentic performance — the largest Qwen ever.

SpaceXAI's frontier 1.5T reasoning model with 500K-token context, agentic coding strength, and real-time X data at roughly half the price of rivals.

Cloudflare's agent-first headless browser that runs entirely in V8 isolates on Workers, using 3-7x less CPU and memory than Chromium for agentic tasks like screenshots and HTML extraction.

Zhipu AI's newest flagship built for coding and cyber defense, squeezing 50% more coding capability out of the GLM-5 base through post-training scaling, ranked #1 among open-source models on Terminal Bench 3.0 and Agents' Last Exam (CLI).

Anthropic's frontier Opus model with 1M-token context and effort-controlled reasoning, near-Fable-5 intelligence at Opus-4.8 pricing for agents and coding.

Meta's coding-focused flagship model powering Muse Code: 1M context, higher first-attempt accuracy, reliable tool calling, and end-to-end developer workflows via the Meta Model API.

SpaceXAI's always-on AI teammates that run on their own cloud computer, sign into your tools, and finish work end-to-end — available in beta for SuperGrok Heavy and Cursor Ultra subscribers.

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.

MiniMax's open-weight music generation model with studio-grade sound, natural vocal synthesis, and full-song generation from prompts and lyrics, with official ComfyUI support.

Free open-source desktop app to run and train AI models 100% locally, with Claude Code and Codex agent integration, media generation, and private web search.

Meta's terminal coding agent powered by Muse Spark 1.2. Install with one command, run persistent background agents across large repos, and pay per token instead of per seat.

Liquid AI's open-weight 2.6B dense model trained for on-device agentic workloads, with 128K context and native tool calling.

Computer use agent that navigates any website end-to-end on a dedicated virtual computer. Book flights, order food, shop, and message candidates on LinkedIn — at a tenth of frontier-model token cost.

PrimeIntellect's self-improving coding agent built on a Recursive Language Model (RLM). A persistent IPython kernel as its single tool, a /refine command that improves the agent itself, and bounded autonomous mode.

Spotify's vendor-neutral agentic development environment. One place to launch, manage, and switch between Claude Code, Gemini CLI, and Codex sessions, with per-session Git worktrees and Portal-powered organizational context.

Open-source terminal and agent manager built for the AI-agent era: organize Claude Code, Codex, and seven more agents into a project tree with live status, split panes, and remote access.

Waku is a fast native macOS app by egoist (ChatWise) that unifies local AI coding agents like Claude Code, Codex, and OpenCode in one Rust + GPUI interface.

Kwaipilot's open-weight agentic coding model: 35B MoE, 3B active per token, Qwen3.6 base, Apache 2.0, 262K context, top PinchBench tool-use score.

Programmable account and virtual wallets for the agentic Internet, combining stablecoin micropayments, delegated identity, and policy-based spending controls.

Open-source agent runtime from Cloudflare that gives every AI agent a durable SQLite-backed filesystem and switchable execution environments (container, isolate shell, or JavaScript isolate) running entirely on Workers.

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

DeepSeek's flagship 1.6T MoE model with 49B active parameters, 1M-token context, MIT open weights, and world-leading coding scores at a fraction of closed-model prices.

DeepSeek's official small agent model with 284B total / 13B active MoE parameters, 1M-token context, MIT open weights, DSpark speculative decoding, and $0.14/$0.28 per 1M tokens pricing.

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.

Alibaba's flagship 2.4T MoE model with 95B active parameters, 1M-token context, native multimodal input, and top-ranked agentic performance — the largest Qwen ever.

OpenAI's frontier flagship with a 1,050K-token context, reasoning effort up to max plus an ultra subagent mode, state-of-the-art Terminal Bench 2.1 coding, and $5/$30 per 1M tokens pricing.

ZCode is Z.ai's official Agentic Development Environment for GLM-5.2: a free desktop app where agents plan, edit, test, and iterate on 1M-token context.

SpaceXAI's frontier 1.5T reasoning model with 500K-token context, agentic coding strength, and real-time X data at roughly half the price of rivals.

Cloudflare's agent-first headless browser that runs entirely in V8 isolates on Workers, using 3-7x less CPU and memory than Chromium for agentic tasks like screenshots and HTML extraction.

Zhipu AI's newest flagship built for coding and cyber defense, squeezing 50% more coding capability out of the GLM-5 base through post-training scaling, ranked #1 among open-source models on Terminal Bench 3.0 and Agents' Last Exam (CLI).

Anthropic's Mythos-class flagship made safe for general use, with 1M-token context, frontier coding and vision, days-long agentic autonomy, and $10/$50 per 1M tokens pricing.

Anthropic's frontier Opus model with 1M-token context and effort-controlled reasoning, near-Fable-5 intelligence at Opus-4.8 pricing for agents and coding.

Moonshot AI's open-weight 2.8T multimodal agentic model with 1M-token context, the world's first open 3T-class model rivaling closed frontier models.

Meta's coding-focused flagship model powering Muse Code: 1M context, higher first-attempt accuracy, reliable tool calling, and end-to-end developer workflows via the Meta Model API.

SpaceXAI's always-on AI teammates that run on their own cloud computer, sign into your tools, and finish work end-to-end — available in beta for SuperGrok Heavy and Cursor Ultra subscribers.

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.

MiniMax's open-weight music generation model with studio-grade sound, natural vocal synthesis, and full-song generation from prompts and lyrics, with official ComfyUI support.

Free open-source desktop app to run and train AI models 100% locally, with Claude Code and Codex agent integration, media generation, and private web search.

Meta's terminal coding agent powered by Muse Spark 1.2. Install with one command, run persistent background agents across large repos, and pay per token instead of per seat.

WeChat AI's 617B MoE model with 23B active parameters, demonstrating the first sequence-length scaling at frontier scale via Hidden Decoding, powering WeChat's Xiaowei assistant.

Liquid AI's open-weight 2.6B dense model trained for on-device agentic workloads, with 128K context and native tool calling.

Open-source durable execution platform for AI agents in TypeScript. Build long-running tasks and chat agents with no timeouts, checkpoint-based resume, retries, queues, and observability — billed only for actual compute.