Hello AppSumo users 👋,
I'm Dhimant, CEO & co-founder at WorkLLM. It's been a few weeks since launch, and I wanted to give you a real update, not just a features list, but where we actually stand and what's shipped.
First, thank you. The feedback from this community has been more direct, detailed, and useful than I expected, and a lot of what's below exists because of specific things you flagged.
New tiers, Tier 5 and Tier 6
For teams that outgrew our original tiers, we've added two higher tiers with more credits and more workspace flexibility. If you've been holding off because your usage didn't fit our original range, these should close that gap.
Credit efficiency
Credit usage was one of the most common pieces of feedback we got, and we took it seriously. We've made several backend improvements, better caching, more efficient memory and document processing, and smarter handling of web search context, all aimed at getting you more out of the same credits. These changes are live and we are expecting 10-20% credit efficiency now.
Full credit transparency
You'll now see exactly what's being used, and when. Credit cost shows up after every answer, every memory update, and every agent test, and there's a live credit meter at the top of your workspace showing your personal usage, workspace-wide usage, and your next refill date, all in one place.
Deep Research + Reasoning controls
You can now choose reasoning depth (Low, Medium, High) alongside deep research, so you're not paying for more thinking than a given task actually needs.
BYOK (Bring Your Own Key)
Now live. Bring your own model provider keys and use them directly inside WorkLLM.
Release notes
We've started publishing proper release notes for every update, so you can see exactly what shipped and when, rather than discovering changes by accident.
Organization memory improvements
You can now upload multiple links at once instead of one at a time, and set links to refresh automatically, daily, weekly, or manually, so your memory stays current without re-adding content by hand.
RTL support
Hebrew and Urdu are now fully supported with correct right-to-left formatting.
Voice input
You can now speak directly into WorkLLM instead of typing.
None of this would have shipped this fast without the specificity of feedback from this community, thread by thread, bug by bug. Keep it coming, it's genuinely shaping the roadmap in real time.
Dhimant
CEO & Co-founder, WorkLLM