{"object":"pro_bench_participant_tracks","schema_version":"pro-bench.participant-tracks/v1","catalog_as_of":"2026-07-23","knowledge_state":"DECLARED","tracks":[{"id":"MODEL_ONLY","label":"Model only","label_ru":"Только модель","description":"One pinned model snapshot, no external tools or product workflow.","description_ru":"Один зафиксированный snapshot модели без внешних инструментов и продуктового workflow.","rankable":true,"required_controls":["pinned_model_snapshot","frozen_prompt_protocol","no_external_tools"]},{"id":"STANDARD_TOOLS","label":"Model + standard tools","label_ru":"Модель + стандартные инструменты","description":"Pinned model snapshots receive the same frozen browser, code and file tools.","description_ru":"Зафиксированные модели получают одинаковый браузер, code runner и файловые инструменты.","rankable":true,"required_controls":["pinned_model_snapshot","identical_tool_contract","frozen_prompt_protocol"]},{"id":"AGENT_SYSTEM","label":"Agent / system","label_ru":"Агент / система","description":"Complete products, coding agents, agent runtimes and orchestrators compete as systems.","description_ru":"Полные продукты, coding agents, agent runtimes и оркестраторы сравниваются как системы.","rankable":true,"required_controls":["product_or_system_snapshot","interaction_protocol","tool_and_model_disclosure"]},{"id":"EQUAL_BUDGET","label":"Best result under equal budget","label_ru":"Лучший результат при равном бюджете","description":"Different participant classes may compete only under the same money, time and operator envelope.","description_ru":"Разные классы участников сравниваются только при одинаковых лимитах денег, времени и оператора.","rankable":true,"required_controls":["shared_money_cap","shared_wall_clock_cap","shared_operator_protocol","actual_cost_and_time_receipts"]}],"participant_classes":["RAW_MODEL","CLOUD_CHAT_PRODUCT","CODING_AGENT","AGENT_RUNTIME","ORCHESTRATOR"],"participants":[{"id":"raw-model-snapshots","display_name":"Pinned API model snapshots","display_name_ru":"Зафиксированные API-модели","vendor":"Multiple providers","participant_class":"RAW_MODEL","catalog_entry_kind":"TEMPLATE","supported_tracks":["MODEL_ONLY","STANDARD_TOOLS","EQUAL_BUDGET"],"access_surfaces":["API"],"model_selection_policy":"FIXED_PINNED","comparison_note":"Every provider model and exact version becomes a separate participant snapshot when admitted by an evidence-bound campaign.","comparison_note_ru":"Каждая модель провайдера и её точная версия становятся отдельным snapshot участника только после evidence-bound кампании.","source_url":"","identity_knowledge_state":"DECLARED","participation_state":"PLANNED","official_score_state":"NOT_EVALUATED"},{"id":"chatgpt-product","display_name":"ChatGPT","vendor":"OpenAI","participant_class":"CLOUD_CHAT_PRODUCT","catalog_entry_kind":"PRODUCT","supported_tracks":["AGENT_SYSTEM","EQUAL_BUDGET"],"access_surfaces":["WEB_APP","DESKTOP_APP","MOBILE_APP"],"model_selection_policy":"USER_OR_PRODUCT_SELECTED","comparison_note":"Evaluate the complete product workflow, including the captured model-selection mode and enabled built-in capabilities; do not relabel it as a raw model run.","comparison_note_ru":"Оценивается полный продуктовый workflow с зафиксированным режимом выбора модели и встроенными возможностями; такой прогон нельзя переименовывать в raw-model.","source_url":"https://help.openai.com/en/articles/20001354-gpt-56-in-chatgpt","identity_knowledge_state":"DECLARED","participation_state":"PLANNED","official_score_state":"NOT_EVALUATED"},{"id":"openai-codex","display_name":"Codex","vendor":"OpenAI","participant_class":"CODING_AGENT","catalog_entry_kind":"PRODUCT","supported_tracks":["AGENT_SYSTEM","EQUAL_BUDGET"],"access_surfaces":["DESKTOP_APP","CLI","IDE","CLOUD_SANDBOX"],"model_selection_policy":"USER_OR_PRODUCT_SELECTED","comparison_note":"Capture the exact Codex surface, sandbox/worktree policy, model configuration and all allowed tools for every campaign snapshot.","comparison_note_ru":"Для каждого snapshot фиксируются поверхность Codex, sandbox/worktree policy, конфигурация модели и все разрешённые инструменты.","source_url":"https://openai.com/codex/","identity_knowledge_state":"DECLARED","participation_state":"PLANNED","official_score_state":"NOT_EVALUATED"},{"id":"anthropic-claude-code","display_name":"Claude Code","vendor":"Anthropic","participant_class":"CODING_AGENT","catalog_entry_kind":"PRODUCT","supported_tracks":["AGENT_SYSTEM","EQUAL_BUDGET"],"access_surfaces":["CLI","WEB_APP","DESKTOP_APP"],"model_selection_policy":"USER_OR_PRODUCT_SELECTED","comparison_note":"Capture the exact execution surface, model configuration, repository permissions and test/tool policy instead of attributing the result to an unspecified Claude model.","comparison_note_ru":"Фиксируются поверхность исполнения, модель, права в репозитории и test/tool policy; результат нельзя приписывать неуказанной модели Claude.","source_url":"https://www.anthropic.com/product/claude-code","identity_knowledge_state":"DECLARED","participation_state":"PLANNED","official_score_state":"NOT_EVALUATED"},{"id":"openclaw","display_name":"OpenClaw","vendor":"OpenClaw Foundation","participant_class":"AGENT_RUNTIME","catalog_entry_kind":"PRODUCT","supported_tracks":["AGENT_SYSTEM","EQUAL_BUDGET"],"access_surfaces":["SELF_HOSTED","MESSAGING_GATEWAY","DEVICE_APPS"],"model_selection_policy":"CONFIGURABLE_PROVIDER","comparison_note":"Pin the OpenClaw release, host configuration, provider/model, skills, memory state and granted device permissions as one participant snapshot.","comparison_note_ru":"В один snapshot участника связываются release OpenClaw, host config, provider/model, skills, начальное состояние памяти и выданные права устройства.","source_url":"https://github.com/openclaw/openclaw","identity_knowledge_state":"DECLARED","participation_state":"PLANNED","official_score_state":"NOT_EVALUATED"},{"id":"nous-hermes-agent","display_name":"Hermes Agent","vendor":"Nous Research","participant_class":"AGENT_RUNTIME","catalog_entry_kind":"PRODUCT","supported_tracks":["AGENT_SYSTEM","EQUAL_BUDGET"],"access_surfaces":["SELF_HOSTED","CLI","MESSAGING_GATEWAY"],"model_selection_policy":"CONFIGURABLE_PROVIDER","comparison_note":"Pin the Hermes Agent release, provider/model, tools, skills and initial memory state; learned skills produced during the run remain part of that run's evidence.","comparison_note_ru":"Фиксируются release Hermes Agent, provider/model, инструменты, skills и начальная память; skills, созданные во время прогона, остаются частью evidence этого прогона.","source_url":"https://github.com/NousResearch/hermes-agent","identity_knowledge_state":"DECLARED","participation_state":"PLANNED","official_score_state":"NOT_EVALUATED"},{"id":"pro-1","display_name":"PRO-1","vendor":"PromanOS","participant_class":"ORCHESTRATOR","catalog_entry_kind":"PRODUCT","supported_tracks":["AGENT_SYSTEM","EQUAL_BUDGET"],"access_surfaces":["API","WEB_APP"],"model_selection_policy":"SYSTEM_ROUTED","comparison_note":"Disclose the exact PRO-1 build, routing policy, selected models, tools, retries, fallbacks, actual cost, wall-clock time and artifact receipts.","comparison_note_ru":"Раскрываются точный build PRO-1, routing policy, выбранные модели, инструменты, retries, fallbacks, фактическая стоимость, wall-clock time и receipts артефактов.","source_url":"/benchmarks/methodology","identity_knowledge_state":"DECLARED","participation_state":"PLANNED","official_score_state":"NOT_EVALUATED"}],"ranking_policy":{"primary_partition_keys":["challenge_version_id","comparison_track","interaction_protocol_id","budget_policy_id"],"cross_track_ranking":"FORBIDDEN","cross_class_comparison":"DESCRIPTIVE_ONLY_UNLESS_EQUAL_BUDGET","equal_budget_ranking_requires":["same_challenge_version","same_money_cap","same_wall_clock_cap","same_operator_protocol","actual_cost_and_time_receipts"]},"participant_snapshot_requirements":["product_or_model_identity","exact_version_or_build","access_surface","account_plan_and_region","model_selection_policy","underlying_model_identity_or_disclosure_state","enabled_tools_connectors_and_permissions","operator_and_interaction_protocol","initial_memory_and_workspace_state","prompt_retries_fallbacks_and_tool_calls","budget_limits_actual_cost_and_wall_clock","outputs_artifacts_hashes_and_receipts"],"score_disclaimer":"Catalog inclusion is planned coverage, not participation, qualification or an official PRO Bench score.","score_disclaimer_ru":"Наличие в каталоге означает планируемое покрытие, а не участие, квалификацию или официальный балл PRO Bench."}