Running OpenClaw with Claude subs is dead. Now what?
by•
As of April 4th, Claude subscriptions no longer cover usage on third-party tools like @OpenClaw.
If you were running your agent with @Claude by Anthropic, curious what option you'd pick: would you pay the extra bundle? pick another provider? or use an alternative like @KiloClaw or @Zo Computer?
1.1K views
Replies
I'm using it via API call still, but just as my orchestrator. The decision from Anthropic finally pushed me to getting a DGX Spark and get more serious about local computer. I use OpenClaw all day for work and personally and I really haven't racked up much of a bill due to spending a ton of time on delegation rules. I still get the big brain of Opus, but i'm using other tools for the muscle. Hope that is helpful.
I think this is where many teams will rediscover that the model choice and the execution environment should be evaluated separately.
Sometimes the cheapest model is not the cheapest workflow once you include retries, weak guardrails, or bad ergonomics.
Would be interesting to hear which alternatives people found good enough in practice, not just on benchmarks.
Hello Aria
This was a rough change for a lot of builders who had workflows running on top of the Claude subscription model. The "free tier via subscription" loophole closing was probably inevitable as these companies figure out API economics.
For what it’s worth, the API pricing for Claude has gotten more accessible over time — running a personal assistant or agentic workflow through the API directly is no longer prohibitively expensive if you’re thoughtful about prompt length and caching.
We run Hello Aria (our AI day-management assistant for WhatsApp/iOS) entirely on API-based LLMs and the unit economics work at scale. Happy to share what we’ve learned about cost optimization if anyone is rebuilding their setup.
I’d probably separate prototyping and production more aggressively now: subscription-friendly tools for exploration, then API or provider redundancy for anything that touches real workflows.
The bigger lesson is not just cost, it’s avoiding single-provider dependency.
This subscription pivot feels like a classic "walled garden" move that nobody invited to the party. While the bundle is convenient, switching to Zo Computer or KiloClaw keeps that independent builder spirit alive. It’s a bit of a headache to migrate, but dodging the "API tax" is usually worth the extra effort. We're all just trying to keep our stacks flexible without getting squeezed by vendor lock-in.
Do you think the convenience of a "one-stop shop" bundle will eventually win over the hassle of managing multiple API keys?