{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:I2OK3RHE4RJIS6EJBVGBVNHGVL","short_pith_number":"pith:I2OK3RHE","schema_version":"1.0","canonical_sha256":"469cadc4e4e4528978890d4c1ab4e6aaf1f92e7749757860de3da95d17104dc4","source":{"kind":"arxiv","id":"2606.20629","version":1},"attestation_state":"computed","paper":{"title":"Specialize Roles, Mix Deployments: Pushing the Cost-Accuracy Frontier of LLM Agent Teams","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.MA","authors_text":"Edoardo Ponti, Liang Cheng, Li Dong, Luo Mai, Wenda Li, Yeqi Huang, Yinsicheng Jiang, Yufan Zhao, Zhan Lu","submitted_at":"2026-05-28T17:12:20Z","abstract_excerpt":"LLM agents are increasingly deployed as multi-role teams, where tasks are divided across specialized roles such as planner, executor, and verifier. In these systems, cost and accuracy are no longer properties of a single model: they depend on which model fills each role and where it is hosted, including API, self-hosted, and hybrid deployment. Existing agentic benchmarks typically evaluate fixed models or fixed agent configurations, and therefore offer limited guidance for cost-accuracy-optimal deployment. We introduce AgentCARD, a role-aware benchmark suite for evaluating LLM agent teams acro"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2606.20629","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.MA","submitted_at":"2026-05-28T17:12:20Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"86a15db728e4f8b837efd5aa649adbd3e3679116db470bfd91a4edbaeb1133fc","abstract_canon_sha256":"9956cd778f561a99108136b8577e7ec8bb4253dfd85d3a6d1587b472bed4dfa8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-23T00:11:51.942595Z","signature_b64":"72JV9r/8hUYNBn42huLr+J/vKUeHMuDe1HketbY9MtAMdrke1WJhllrillr8FbmZaMhO+3szUTlQ6/KijhJAAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"469cadc4e4e4528978890d4c1ab4e6aaf1f92e7749757860de3da95d17104dc4","last_reissued_at":"2026-06-23T00:11:51.942184Z","signature_status":"signed_v1","first_computed_at":"2026-06-23T00:11:51.942184Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Specialize Roles, Mix Deployments: Pushing the Cost-Accuracy Frontier of LLM Agent Teams","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.MA","authors_text":"Edoardo Ponti, Liang Cheng, Li Dong, Luo Mai, Wenda Li, Yeqi Huang, Yinsicheng Jiang, Yufan Zhao, Zhan Lu","submitted_at":"2026-05-28T17:12:20Z","abstract_excerpt":"LLM agents are increasingly deployed as multi-role teams, where tasks are divided across specialized roles such as planner, executor, and verifier. In these systems, cost and accuracy are no longer properties of a single model: they depend on which model fills each role and where it is hosted, including API, self-hosted, and hybrid deployment. Existing agentic benchmarks typically evaluate fixed models or fixed agent configurations, and therefore offer limited guidance for cost-accuracy-optimal deployment. We introduce AgentCARD, a role-aware benchmark suite for evaluating LLM agent teams acro"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.20629","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2606.20629/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2606.20629","created_at":"2026-06-23T00:11:51.942245+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.20629v1","created_at":"2026-06-23T00:11:51.942245+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.20629","created_at":"2026-06-23T00:11:51.942245+00:00"},{"alias_kind":"pith_short_12","alias_value":"I2OK3RHE4RJI","created_at":"2026-06-23T00:11:51.942245+00:00"},{"alias_kind":"pith_short_16","alias_value":"I2OK3RHE4RJIS6EJ","created_at":"2026-06-23T00:11:51.942245+00:00"},{"alias_kind":"pith_short_8","alias_value":"I2OK3RHE","created_at":"2026-06-23T00:11:51.942245+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/I2OK3RHE4RJIS6EJBVGBVNHGVL","json":"https://pith.science/pith/I2OK3RHE4RJIS6EJBVGBVNHGVL.json","graph_json":"https://pith.science/api/pith-number/I2OK3RHE4RJIS6EJBVGBVNHGVL/graph.json","events_json":"https://pith.science/api/pith-number/I2OK3RHE4RJIS6EJBVGBVNHGVL/events.json","paper":"https://pith.science/paper/I2OK3RHE"},"agent_actions":{"view_html":"https://pith.science/pith/I2OK3RHE4RJIS6EJBVGBVNHGVL","download_json":"https://pith.science/pith/I2OK3RHE4RJIS6EJBVGBVNHGVL.json","view_paper":"https://pith.science/paper/I2OK3RHE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.20629&json=true","fetch_graph":"https://pith.science/api/pith-number/I2OK3RHE4RJIS6EJBVGBVNHGVL/graph.json","fetch_events":"https://pith.science/api/pith-number/I2OK3RHE4RJIS6EJBVGBVNHGVL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/I2OK3RHE4RJIS6EJBVGBVNHGVL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/I2OK3RHE4RJIS6EJBVGBVNHGVL/action/storage_attestation","attest_author":"https://pith.science/pith/I2OK3RHE4RJIS6EJBVGBVNHGVL/action/author_attestation","sign_citation":"https://pith.science/pith/I2OK3RHE4RJIS6EJBVGBVNHGVL/action/citation_signature","submit_replication":"https://pith.science/pith/I2OK3RHE4RJIS6EJBVGBVNHGVL/action/replication_record"}},"created_at":"2026-06-23T00:11:51.942245+00:00","updated_at":"2026-06-23T00:11:51.942245+00:00"}