Warp tested every model on its own tasks: Anthropic great but pricey
thursdai_pod · x · 2026-11-18
Warp founder Zach Lloyd said the team "just tested everything" — models and harnesses — on Warp's own tasks. His take: "The Anthropic models are awesome, but we do find they're pretty expensive."
More from coding & agent
- Non-engineer ships real SaaS via Replit vibe coding for just $541 — socialwithaayan · 2026-11-18
- Hands-on: GPT-6.2 Astra & Sol shine at computer use and software testing — craigsdennis · 2026-11-18
- Together AI talks with Ricursive Intelligence on building a frontier local coding model — togethercompute · 2026-11-18
- Agent kept bending arms wrong, so this dev wired MuJoCo into an MCP server — usefulwrong · 2026-11-18
- Google rebrands Spark AI agent as 'Tasks,' now free, can call businesses — The Verge AI · 2026-11-18
- YC demo builds a movie inside a meeting: speech-to-video agent streams in real time — ycombinator · 2026-11-18