AI tools,
put to work.

Explore unfamiliar code, work out how to check a change and make the edit with help from OpenAI’s terminal coding assistant.

Follow a failure through the source and tests, then work on a fix with Anthropic’s coding assistant in your terminal.

Choose the model and adapt the assistant around recurring work, such as turning incident files into a report another engineer can check.

See which coding session needs you and return to its conversation, with terminal processes that keep running while you detach.

Get another review of a pull request, discuss findings beside the code and check whether the suggested problems can actually occur.

Turn a written report into a defined choice or score, then compare its recommendations with the decisions your workflow needs.

Manus suits a task whose answer needs to become a deliverable: a source-linked briefing, slides or a small website. Agent mode plans and performs multi-step work, while Chat mode offers a lighter conversation.

Gemini Notebook helps a reader understand a selected body of evidence rather than start with an unrestricted web question. Google's former NotebookLM URL now redirects to the current app, and NotebookLM remains a useful search alias.

Perplexity is a starting point for a question that needs current web evidence and citations the reader can inspect. Search and deeper research can bring several sources into an answer; file analysis and model selection depend on the plan.

Notion AI is useful when the question and its supporting material already live in a team's Notion workspace. It can find and transform that knowledge, search connected apps and the web, draft or edit documents, analyze files and populate database summaries.

Cursor brings coding assistance into work on an existing repository through editor, CLI and cloud-agent surfaces. Agent can search files, edit code and run commands; Ask provides a read-only exploration option.

Replit Agent is a build-and-deploy environment for someone who wants to make an app without first preparing a local development setup. Describe a small application, review the plan and iterate in the project editor before deliberately publishing.

n8n combines explicit workflow logic with model-driven steps, so a repeatable process can reserve AI for the part that needs interpretation. It is available as managed Cloud and self-hosted software under fair-code terms that need checking for the intended use.

Make AI Agent (New) places agent work inside a cloud automation scenario. Configure instructions, a provider, knowledge, inputs and tools, then test the agent as part of that scenario.

Zapier is moving its standalone Agents product into AI by Zapier steps in the Zap editor. As of October 2, 2026, no Agents shutdown date has been announced.

Lovable develops web apps from a natural-language description into editable code, backend behavior and a deployment workflow. It suits a small app whose owner wants to iterate in a preview before deciding what should be published.

Higgsfield puts image and video generation, reference-based editing and preset-led creative work in one browser platform. It combines its own model families, including Soul, with separately identified partner models such as Kling and Veo; choosing the platform does not choose a single generation engine.

Runway offers direct image/video tools, task-specific Apps, a conversational creative Agent and node-based Workflows. These are different ways to organize a production job rather than interchangeable model names.

Firefly is both Adobe’s generative model family and an application that also offers partner models. An asset made in the application therefore needs a model record before Adobe-model training claims can be attached to it.

Midjourney generates and edits images and can animate a starting image into video. It is suited to exploring a visual direction before committing to a finished asset: composition, lighting and style can be compared against one brief, then a selected draft can be refined.

ElevenLabs supplies speech-generation tools through a web creative interface and APIs. Keep three choices distinct: the product used for the task, the model producing speech and the voice representing the speaker.

Descript edits recorded audio and video through a transcript linked to the media, alongside timeline controls, scenes and captions. That makes it a different starting point from generating a new shot: import an authorized recording, correct its transcript, then remove or rearrange words while checking the resulting cuts.

Canva AI works inside Canva’s Visual Suite to turn a brief into a design that can be edited further. Its Design Model announcement describes layered output, while the live product/help pages retain account-dependent rollout and allowances; not every account has identical AI 2.0 features.

Suno creates songs or instrumental sketches from a description, with web and mobile access. In Simple Mode, a prompt describes the desired sound and an Instrumental toggle excludes requested vocals.