{"paper":{"title":"Point-E: A System for Generating 3D Point Clouds from Complex Prompts","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A two-stage diffusion process turns text prompts into 3D point clouds in 1-2 minutes on one GPU.","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Alex Nichol, Heewoo Jun, Mark Chen, Pamela Mishkin, Prafulla Dhariwal","submitted_at":"2022-12-16T23:22:59Z","abstract_excerpt":"While recent work on text-conditional 3D object generation has shown promising results, the state-of-the-art methods typically require multiple GPU-hours to produce a single sample. This is in stark contrast to state-of-the-art generative image models, which produce samples in a number of seconds or minutes. In this paper, we explore an alternative method for 3D object generation which produces 3D models in only 1-2 minutes on a single GPU. Our method first generates a single synthetic view using a text-to-image diffusion model, and then produces a 3D point cloud using a second diffusion model"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Our method first generates a single synthetic view using a text-to-image diffusion model, and then produces a 3D point cloud using a second diffusion model which conditions on the generated image. ... produces 3D models in only 1-2 minutes on a single GPU.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That a single synthetic 2D view generated by the text-to-image model contains enough information for the second diffusion model to recover accurate 3D geometry for complex prompts.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Point-E is a cascaded diffusion system that generates 3D point clouds from text in minutes by first synthesizing a 2D view then lifting it to 3D.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A two-stage diffusion process turns text prompts into 3D point clouds in 1-2 minutes on one GPU.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"5e39034ff84f2df9fa7168a4bd601b781e54144619976805ae4c298c916d599e"},"source":{"id":"2212.08751","kind":"arxiv","version":1},"verdict":{"id":"b381caa9-b0e7-4fe5-b287-6b08e21df986","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T20:46:07.223132Z","strongest_claim":"Our method first generates a single synthetic view using a text-to-image diffusion model, and then produces a 3D point cloud using a second diffusion model which conditions on the generated image. ... produces 3D models in only 1-2 minutes on a single GPU.","one_line_summary":"Point-E is a cascaded diffusion system that generates 3D point clouds from text in minutes by first synthesizing a 2D view then lifting it to 3D.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That a single synthetic 2D view generated by the text-to-image model contains enough information for the second diffusion model to recover accurate 3D geometry for complex prompts.","pith_extraction_headline":"A two-stage diffusion process turns text prompts into 3D point clouds in 1-2 minutes on one GPU."},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"3a6f18a3cc732cb7fe6b56166da9f2c8e20a96c8d506859fec384aa79bd604ba"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}