The pace of AI tooling in 2025–2026 has been genuinely hard to keep up with. Every few weeks there’s a new model, a new framework, a new capability that renders some part of your workflow obsolete — or dramatically better.

I wanted to open a thread for something simple: what’s actually working for you right now?

Not what you’ve read about. Not what’s in the press. What tools are you actively using, week in week out, that earn their place in your stack?

—

I’ll go first. My current rotation:

For writing and thinking:
Claude — specifically for longer-form work and anything requiring nuance. The context window and instruction-following are still the best in class for my use cases.

For images:
Midjourney (v6/v7) for anything stylistic or conceptual. DALL-E 4 for precise text-in-image tasks and when I need to iterate fast.

For video:
RunwayML Gen-4 for polished short-form clips. Still experimenting with Kling for more dynamic motion.

For voice:
ElevenLabs for narration. The new voice cloning fidelity is remarkable — I’ve been building audio content that would have taken a full production setup two years ago.

For agents:
OpenClaw for my personal assistant layer. It connects my chat apps, handles background tasks, and keeps persistent memory across sessions. It’s become infrastructure at this point — I’d notice its absence more than most apps I pay for.

—

This community feels like exactly the right place to compare notes on all of this. The overlap between tools here — prompting, video, voice, agents — is where the interesting workflows live.

What’s in your current stack? What’s earned a permanent spot, and what are you still evaluating?