We’re hosting the world's first Autonomous Company Hackathon @ycombinator $100,000+ in prizes. Winners get a guaranteed YC interview. Mac Minis, iPhone 18 + much more: 💰💰 Pick a business. Build the agent team that runs it with the Naïve SDK. Show it operating at demo time. November 7–8, YC office, San Francisco. Thanks to our sponsors: @OpenAI , @OpenRouter , @supabase, @neondatabase, @vercel , @stripe , @agentmail, @tenki, @MonidHQ, @apify, and more. register ⬇️

Introducing Clipping Bench v1.0. We measure how well agents can run profitable clipping channels autonomously. To create this benchmark, we gave each team of agents $300 to run hundreds of channels across Tiktok, Instagram, Youtube (120M+ views) using SOTA models.

Takeaways: > Baseline strategy*: $2,606 > Best model, GPT-6 Astra: $9,514 > Oracle**: $27,521 (2.9x Astra) *Baseline strategy: a scripted bot, no AI model, that reads transcripts and follows the posted campaign rules. **Oracle: a scripted bot that can see the simulator's hidden state. It shows roughly how much a near-perfect player could earn. Conclusion: There's a lot of room left for improvement for agents to autonomously run companies.

Announcing Naïve × @tenki Starting today, every Naïve agent computer runs on Tenki sandboxes. On average, users will see: → ~30% lower sandbox costs → 72.3% lower sandbox latency

Introducing Vetta, the most efficient harness for long-horizon agent tasks. Same model, same tasks. Only the harness changed: 🥇 Vetta: $0.298 per finished task 🥈 claude-code: $0.872 🥉 hermes: $1.095 Highest finish rate, a third of the price. Autonomous companies only happen when the cost makes sense. Here's how we get there:

Naïve Lab is live four benchmarks on the substrate agents actually run on. Serverless agent runtime hosting: what a million agents cost to keep alive, at 100,000 tenants × 10 agents per month. 🥇 Naïve · serverless — ~$44–60k 🥈 Self-host Hermes / Eve — ~$740k 🥉 Self-host OpenClaw / Paperclip — ~$1.5M VM per tenant (AWS) — ~$14M Every runtime here is free to licence. The cost is the box staying awake between turns, and agent work is bursty — a turn, then nothing for an hour. We want an agent running continuously for every workspace. That is a hosting problem before it is a model problem.