{"type":"rich","version":"1.0","provider_name":"Transistor","provider_url":"https://transistor.fm","author_name":"START","title":"🎧 START pod: Tejas Bhakta, Founder & CEO , Morph \"Subagents and tools that improve coding agents\"","html":"<iframe width=\"100%\" height=\"180\" frameborder=\"no\" scrolling=\"no\" seamless src=\"https://share.transistor.fm/e/9464f985\"></iframe>","width":"100%","height":180,"duration":1215,"description":"Agents don’t need bigger models. They need better tools.\nMorph trains coding subagents.\nNot for humans. For frontier models.\n\nFast Apply edits at 10,000 tokens/sec.\nWarpGrep handles code and log search.\n\nBoth keep the main model’s context clean\nBecause when context gets too large, performance drops.\n\nNow Morph is pushing coding subagents even faster.\nOne newer model runs at 33,000 tokens/sec: https://docs.morphllm.com/sdk/components/compact\n\n🎙️ Tejas Bhakta, Founder & CEO, Morph\n01:30 Fast Apply + WarpGrep\n02:26 Context fills up around 100k\n02:38 Keep the main model context clean\n03:31 “You can’t scale human attention 100x”\n07:28 Founders are missing how to value human attention\n08:30 New model at 33,000 tokens/sec\n10:08 Better, faster, and cheaper than the frontier","thumbnail_url":"https://img.transistorcdn.com/q-E1hh7K6IS4AfZiNy2p4MYVGcUOO8lQP92h8QbOEOA/rs:fill:0:0:1/w:400/h:400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS80NzVj/MDEzNjkxNjU1N2Uy/NDFhMDQ3M2ZhNWI3/NWY0MS5wbmc.webp","thumbnail_width":300,"thumbnail_height":300}