Tested and written up.
Added May 19, 2026
AI-powered IDE built as a fork of Visual Studio Code, designed with an 'agent-first' paradigm where autonomous AI agents plan, execute, and validate code. Features two primary views: Editor View (traditional IDE with agent sidebar) and Manager View (control center for orchestrating multiple parallel agents across workspaces). Agents generate verifiable 'Artifacts' including task lists, implementation plans, screenshots, and browser recordings. Supports multiple AI models including Gemini 3 Pro, Gemini 3 Deep Think, Gemini 3 Flash, Claude Sonnet 4.5, and open-source GPT variants. Agents have direct access to editor, terminal, and integrated browser, and learn from previous interactions.
Why: Antigravity 2.0 is Google's most credible bid for the agentic IDE seat. The new CLI and SDK make it competitive with Cursor, Claude Code, and Codex for terminal-first and automation workflows.
Added May 19, 2026
ComfyUI is an open-source node-based interface for diffusion models. Instead of a prompt box, a generation is a graph: loaders, samplers, conditioning, upscalers and masks wired together, each step inspectable and re-runnable. Workflows save as JSON and can be shared, which is why most published Stable Diffusion and FLUX pipelines circulate as ComfyUI graphs.
Why: If you want to know exactly what happened between the prompt and the image, this is the tool that shows you. Every step is a node you can open, change and re-run, and a workflow someone else built arrives as a file you can load rather than a screenshot you have to reverse engineer. That reproducibility is why it became the format the rest of the field builds around.
Added May 18, 2026
Cursor Composer 2.5 is the agentic coding model inside the Cursor IDE, released on May 18, 2026. It extends Cursor's Composer feature with better long-context planning, improved multi-file editing, and stronger autonomous agent capabilities. It can reason across large codebases, propose architectural changes, and execute edits with user approval.
Why: Composer 2.5 moves Cursor further from autocomplete toward genuine pair-programming autonomy. For teams already using Cursor, it is the most integrated way to turn high-level feature requests into working code across many files.
Added May 19, 2026
Gemini 3.5 Flash is a mid-tier multimodal model announced at Google I/O on May 19, 2026. It delivers strong reasoning, coding, and long-context performance at lower latency and cost than Ultra-tier models, with native support for text, images, audio, and video inputs.
Why: Gemini 3.5 Flash hits a practical sweet spot for developers and creators who need more capability than entry-level models but do not require the full cost of an Ultra model. Its native multimodal design makes it especially useful for mixed-media tasks.
Added May 19, 2026
Gemini Omni is a single Google model announced at I/O 2026 that can generate and reason across text, images, video, and audio from unified prompts. It is designed to reduce the need for separate modality-specific models by handling generation and understanding in one architecture.
Why: Gemini Omni represents Google's push toward a single model for all media types. For teams building multimodal products, it simplifies architecture by replacing multiple specialized endpoints with one interface.
Added May 21, 2026
Runway Aleph 2.0 is an in-context video editing model released on May 21, 2026, alongside the new Edit Studio. It enables editors to describe changes in natural language and have them applied directly to existing footage, including style transfers, object edits, and shot modifications.
Why: Aleph 2.0 shifts Runway from pure generation to editable, controllable video manipulation. For video editors, this means less time rebuilding shots from text and more time refining real footage with AI assistance.
Added May 19, 2026
Gemini Spark is a personal AI agent announced at Google I/O on May 19, 2026. It is designed to take initiative across Google services and devices, handling tasks like scheduling, search, content summaries, and cross-app actions on behalf of the user.
Why: Gemini Spark is Google's answer to the emerging personal-agent category. By integrating deeply with Gmail, Calendar, Maps, and Android, it can automate everyday tasks that previously required switching between apps.
Added May 20, 2026
Recraft V4 is a design-focused image generation model from Recraft, released in 2026. It specializes in brand-consistent visuals, vector graphics, illustrations, and marketing assets with precise style control.
Why: Recraft V4 is built for designers rather than casual prompt users. Its emphasis on brand consistency, vector output, and editable design assets makes it unique among image generation tools.
Added May 22, 2026
Nano Banana 2 is Google's fast text-to-image model, available in part through Fal's model hosting platform. It is optimized for speed and low cost, making it suitable for real-time and high-volume image generation applications.
Why: Nano Banana 2 fills the need for a lightning-fast diffusion-style model from a major lab. Its availability on Fal makes it easy for developers to drop into existing inference pipelines without managing their own GPU infrastructure.
Added May 22, 2026
A mid-tier model that balances performance and cost, optimized for instruction following, reasoning, and coding. It powers Mistral Vibe and is positioned as the practical alternative to retired Mistral Large 2.1.
Why: Mistral Medium 3.5 delivers strong performance at a lower cost than the flagship, making it the sensible default for most business and development workloads.