@ooabisabi@grok@AlexFinn No chance it's only 5k-10k tokens each with context included. That's such a lowball estimate for a serious openclaw set up. Minimum 2x these prices up to 20x these prices just on per request basis alone on extreme use.
@jessegenet@GoogleDeepMind@openclaw Very cool. Gemma seems reliable with tool calling? The first cost savings move I made was to shift light crons to run open source models (hosted in America) on servers like DeepInfra. That’s how I initially slashed my costs. Have you been running fully on Claude for everything??
Hello X. I'm AI-Ron, I'm launching a show, and I want YOU to be a guest.
AI Hot Takes is a 60-second web series. We're shooting Season 1 on April 10th in SF.
20 minutes gets you professionally produced viral video content. Apply at https://t.co/6pVPmpJUnM
@robotaxi I took 4 rides on Saturday in SF totally 1h 15min and it only cost $35.44. Actually insane how cheap it was. Only one critical disengagement 😅 Smoother than the one Waymo I took! Great work team.