Hey Indie Hackers — I’m Samson from India, building under Acsaven.
I was juggling OpenAI / Anthropic / Gemini / OpenRouter spend and finding out the hard way at invoice time. I wanted one place to sync, a limit I set, and an email when I crossed it.
So I shipped TryTokka:
- Multi-provider sync
- Scout emails after daily sync when you cross your limit
- Sync now for same-day numbers
- Spike investigation by model + savings hints
- Read-only keys, AES-256
Pricing: free forever (limited) · Solo $9 · Team $29 · 7-day trial, no card.
What I want feedback on:
1) What threshold do you actually set (monthly $ vs % of budget)?
2) Which provider should I prioritize next?
Not TikTok. https://trytokka.com/
What I like here is that you're shifting spend from something people discover in hindsight to something they can respond to while decisions are still reversible.
That changes the product from reporting costs to influencing behavior, even if the underlying data stays exactly the same.
That's exactly the shift I was going for — most spend tools just show you a bill after the fact, which is basically an autopsy. I wanted something that catches you mid-decision instead. Appreciate you putting it that way — might borrow that framing for the landing page.
I'm glad it resonated.
I'd be interested in continuing the conversation by email if you're open to it. What's the best email to reach you on?
That's fair.
I'd rather not unpack my perspective over a public thread because I think it deserves a longer conversation than comments allow.
If you're open to it, what's the best email to reach you on?
Avoid invoice shock’ is a strong pain point. Are users mostly individuals or teams — and how early in the spend cycle are they engaging?
That ‘second provider = real problem’ signal is actually a really strong implicit qualifier.
The interesting opportunity there is catching people right before the surprise invoice, not after.
Have you considered any early warning signals (usage spikes, pattern changes) that could trigger value before the pain hits?
Yes, actually — that's already shipped, not just an idea. On Solo/Team there's forecasting and anomaly detection running alongside the flat $ threshold, plus spike investigation broken down by model. So it's not purely "you set a number, we wait for you to cross it" — it can flag an unusual burn pattern before you'd naturally hit that monthly limit. The threshold alert is the floor; the spike detection is the part meant to catch it earlier than that.
That’s a strong setup — spike detection is where the real value is.
The interesting angle might be how early you can surface “this will become a problem” rather than “this is a problem.”
Are you seeing more solo devs or teams hitting those anomalies first?
Quick add: Scout emails after daily sync when you cross the limit you set — or use Sync now for same-day numbers. Read-only keys; we’re not a proxy and not TikTok.
Curious what monthly threshold you’d set first if you connect two providers tomorrow.