--- license: cc-by-sa-4.0 language: - eng language_details: >- en tags: - task-completion - agents - llm-agents - dialogue-system - task-oriented-dialogue-system - user-persona - user-simulator - role-playing task_categories: - text-generation configs: - config_name: all data_files: - split: test path: output_jsonl/*.jsonl - config_name: multi_8domain data_files: - split: test path: output_jsonl/multi_8domain.jsonl - config_name: multi_11domain data_files: - split: test path: output_jsonl/multi_11domain.jsonl - config_name: attraction data_files: - split: test path: output_jsonl/attraction.jsonl - config_name: bar data_files: - split: test path: output_jsonl/bar.jsonl - config_name: cafe data_files: - split: test path: output_jsonl/cafe.jsonl - config_name: cruise data_files: - split: test path: output_jsonl/cruise.jsonl - config_name: dessert data_files: - split: test path: output_jsonl/dessert.jsonl - config_name: flight data_files: - split: test path: output_jsonl/flight.jsonl - config_name: food&dining data_files: - split: test path: output_jsonl/food&dining.jsonl - config_name: hotel data_files: - split: test path: output_jsonl/hotel.jsonl - config_name: kayak data_files: - split: test path: output_jsonl/kayak.jsonl - config_name: liveshow data_files: - split: test path: output_jsonl/liveshow.jsonl - config_name: vehicle data_files: - split: test path: output_jsonl/vehicle.jsonl - config_name: multi_2domain_attraction_cafe data_files: - split: test path: output_jsonl/multi_2domain_attraction_cafe.jsonl - config_name: multi_2domain_attraction_dessert data_files: - split: test path: output_jsonl/multi_2domain_attraction_dessert.jsonl - config_name: multi_2domain_food&dining_hotel data_files: - split: test path: output_jsonl/multi_2domain_food&dining_hotel.jsonl - config_name: multi_2domain_hotel_vehicle data_files: - split: test path: output_jsonl/multi_2domain_hotel_vehicle.jsonl - config_name: multi_2domain_liveshow_food&dining data_files: - split: test path: output_jsonl/multi_2domain_liveshow_food&dining.jsonl - config_name: multi_2domain_vehicle_attraction data_files: - split: test path: output_jsonl/multi_2domain_vehicle_attraction.jsonl - config_name: multi_3domain_attraction_food&dining_cafe data_files: - split: test path: output_jsonl/multi_3domain_attraction_food&dining_cafe.jsonl - config_name: multi_3domain_flight_hotel_bar data_files: - split: test path: output_jsonl/multi_3domain_flight_hotel_bar.jsonl - config_name: multi_3domain_hotel_attraction_vehicle data_files: - split: test path: output_jsonl/multi_3domain_hotel_attraction_vehicle.jsonl - config_name: multi_3domain_vehicle_food&dining_hotel data_files: - split: test path: output_jsonl/multi_3domain_vehicle_food&dining_hotel.jsonl - config_name: multi_4domain_flight_hotel_attraction_vehicle data_files: - split: test path: output_jsonl/multi_4domain_flight_hotel_attraction_vehicle.jsonl - config_name: multi_4domain_hotel_vehicle_food&dining_bar data_files: - split: test path: output_jsonl/multi_4domain_hotel_vehicle_food&dining_bar.jsonl --- # T1-Bench: Benchmarking Multi-Scenario Agents in Real-World Domains **T1-Bench** is a high-fidelity benchmark for evaluating task-completion and role-playing agents across **25 domains**, including **11 single-domain and 14 multi-domain settings**. It provides **76 tools** and **extensive human annotations**, enabling systematic evaluation of agents in realistic, policy-grounded multi-domain interactions with natural user–assistant role-playing. ![Overview](./assets/architecture.png) **T1-Bench** is a fully automated benchmark for evaluating the tool-calling capabilities of conversational AI agents across diverse service domains in task-oriented settings. The framework simulates end-to-end user–agent interactions without requiring human intervention at inference time, where a User Agent generates realistic customer utterances conditioned on predefined task goals and a tool-augmented Assistant Agent responds by invoking domain-specific tools/APIs and producing outputs grounded in tool results, prior conversational context, and domain-specific policies. Designed to capture the sequential and interactive nature of real-world service workflows across multi-domain scenarios, T1-Bench requires agents to maintain conversational state, reason over prior tool outputs, and execute multi-step operations such as search, filtering, booking, modification, and cancellation. Because all interactions are grounded in deterministic datasets and executable tools, the benchmark enables reproducible and fine-grained evaluation of agent behavior, tool-use decisions, and task completion performance. The paper is accepted at EMNLP Findings 2026. Check our [ArXiv](https://arxiv.org/abs/2606.11070). ## Dataset structure Each row has the following columns: - `domain_id`: the domain name concatenated with a row index (e.g. `attraction-0`). - `domain`: the domain name, matching the source folder name. - `user_template`: the persona, user profile, and goal used to drive the user simulator. - `benchmark_success_criteria`: a JSON-stringified object describing the exact-match tool calls and LLM-as-judge metrics used to score the conversation. ## Domains 11 single-domain configs: `attraction`, `bar`, `cafe`, `cruise`, `dessert`, `flight`, `food&dining`, `hotel`, `kayak`, `liveshow`, `vehicle`. 14 multi-domain configs, prefixed with `multi_Xdomain_` where `X` is the number of entities combined: `multi_2domain_attraction_cafe`, `multi_2domain_attraction_dessert`, `multi_2domain_food&dining_hotel`, `multi_2domain_hotel_vehicle`, `multi_2domain_liveshow_food&dining`, `multi_2domain_vehicle_attraction`, `multi_3domain_attraction_food&dining_cafe`, `multi_3domain_flight_hotel_bar`, `multi_3domain_hotel_attraction_vehicle`, `multi_3domain_vehicle_food&dining_hotel`, `multi_4domain_flight_hotel_attraction_vehicle`, `multi_4domain_hotel_vehicle_food&dining_bar`, `multi_8domain`, `multi_11domain`. ## Results ![Results](./assets/pass_at_k3_chart.png) ![PassK](./assets/complexity_plot_pass_k.png) ## Code Please refer to the official implementation repository: [https://github.com/CapitalOne-Research/t1-bench](https://github.com/CapitalOne-Research/t1-bench) ## Contact E-mail: [Genta Indra Winata](mailto:genta.winata@capitalone.com) or [Amartya Chakraborty](mailto:amartya.chakraborty@capitalone.com). ## License The dataset is licensed under CC-BY-SA 4.0. ## Citation If you find this dataset useful, please cite the following work ```bibtex @article{winata2026t1, title={T1-Bench: Benchmarking Multi-Scenario Agents in Real-World Domains}, author={Winata, Genta Indra and Chakraborty, Amartya and Lin, Yuzhen and Rao, Swasthi P and Siingh, Shikhhar and Lu, Houhan and Bathaee, Nadia and Hatwar, Sriharsha and Dashore, Paresh and Jain, Anmol and others}, journal={arXiv preprint arXiv:2606.11070}, year={2026} } ```