{"id":"bench-tau-bench-bench--taubench-com","name":"τ-bench (tau-bench / τ²-bench)","url":"https://taubench.com/","page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/bench-tau-bench-bench--taubench-com","family":"memory-research","section":"Memory benchmarks & evaluation","sections":["Memory benchmarks & evaluation"],"tier":3,"last_verified_at":"2026-10-29","lifecycle":{"status":"active","severity":0,"note":null},"taxonomy":{"storage":[{"value":"n/a","primary":true,"reason":"not applicable — eval dataset, not a system"}],"retrieval":[{"value":"n/a","primary":true,"reason":"not applicable — eval dataset, not a system"}],"persistence":[{"value":"n/a","primary":true,"reason":"not applicable — eval dataset, not a system"}],"update":[{"value":"read-only","primary":true,"reason":null}],"unit":[{"value":"n/a","primary":true,"reason":"not applicable — eval dataset, not a system"}],"governance":[{"value":"n/a","primary":true,"reason":"not applicable — eval dataset, not a system"}],"conflict":[{"value":"n/a","primary":true,"reason":"not applicable — read-only evaluation dataset"}]},"cells":{"type":{"value":"Benchmark","citation":"https://taubench.com/","status":"real-data","tier":"T2","source_class":"vendor"},"desc":{"value":"Sierra's benchmark of whether agents can converse with simulated users, call tools and follow policy in domains such as retail, airline, telecom and banking. τ²-bench adds dual-control environments, and later releases add voice and knowledge tasks.","citation":"https://taubench.com/","status":"real-data","tier":"T2","source_class":"vendor"},"claims":{"value":"Abstract: Existing benchmarks do not test language agents on their interaction with human users or ability to follow domain-specific rules, both of which are vital for deploying them in real world applications. We propose $\\tau$-bench, a benchmark emulating dynamic conversations between a user (simulated by language models) and a language agent provided with domain-specific API tools and policy guidelines. We employ an efficient and faithful evaluation process that compares the database state at the end of a conversation with the annotated goal state. We also propose a new metric (pass^k) to evaluate the reliability of agent behavior over multiple trials. Our experiments show that even state-of-the-art function calling agents (like gpt-4o) succeed on &lt;50% of the tasks, and are quite inconsistent (pass^8 &lt;25% in retail). Our findings point to the need for methods that can improve the ability of agents to act consistently and follow rules reliably.","citation":"https://arxiv.org/abs/2406.12045","status":"real-data","tier":"T2","source_class":"paper"},"modalities":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"created":{"value":"2025-06-09","citation":"https://github.com/sierra-research/tau2-bench","status":"real-data","tier":"T1","source_class":"repo"},"latest-release":{"value":"last commit 2026-10-29","citation":"https://github.com/sierra-research/tau2-bench","status":"real-data","tier":"T1","source_class":"repo"},"license":{"value":"MIT","citation":"https://taubench.com/","status":"real-data","tier":"T2","source_class":"vendor"},"gh":{"value":"https://github.com/sierra-research/tau2-bench (2354★)","citation":"https://github.com/sierra-research/tau2-bench","status":"real-data","tier":"T1","source_class":"repo"},"mindshare":{"value":"2354 GitHub stars","citation":"https://github.com/sierra-research/tau2-bench","status":"real-data","tier":"T1","source_class":"repo"},"citations":{"value":"searched not found","citation":"https://arxiv.org/abs/2406.12045","status":"depth-floor-reached","tier":"T2","source_class":"paper"},"funding":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"customers":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"pricing":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"compliance":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"data-handling":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"deployment":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"hq":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"founders":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"volume":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"perf":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"repro":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"code-release":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"api-surface":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"latency":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"throughput":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"backend-storage":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"multi-tenancy":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"encryption":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"sso-rbac":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"embedding-model":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"consistency":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"versioning":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"tombstoning":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"schema-evolution":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"namespace":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"contradiction":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"forgetting":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"mcp-support":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"a2a-support":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"otel":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"webhooks":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"import-export":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"integration-count":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"orchestration":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"programmatic-control":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"vendor-benchmarks":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"pricing-specifics":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"agent-abstraction":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"memory-primitives":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"llm-lock":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"runtimes":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"session-handling":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"validated-verticals":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"time-to-running":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"anti-fit":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"optimised-for":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"adjacent-infrastructure":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"pros":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"cons":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"links":{"value":"https://taubench.com/ ; https://github.com/sierra-research/tau2-bench ; https://arxiv.org/abs/2406.12045","citation":"https://taubench.com/","status":"real-data","tier":"T2","source_class":"vendor"},"obs-langsmith":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"obs-opentelemetry":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"obs-datadog":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"obs-helicone":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"obs-weave":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"obs-langfuse":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"obs-arize":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"obs-custom":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"cost-token-budget":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"cost-prompt-caching":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"cost-semantic-caching":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"cost-batching":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"cost-model-routing":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"cost-streaming-only":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"cost-observability-cost-attribution":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"eval-langsmith-evals":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"eval-braintrust":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"eval-weights-and-biases-agent":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"eval-helicone-evals":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"eval-custom-test-harness":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"eval-human-loop":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"eval-production-traffic-replay":{"value":"no","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"commit-trajectory":{"value":"searched not found","citation":"https://github.com/sierra-research/tau2-bench","status":"depth-floor-reached","tier":"T1","source_class":"repo"},"citation-trajectory":{"value":"not applicable — non-paper row","citation":null,"status":"not-applicable","tier":"T3"},"download-trajectory":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"cost-input-usd-per-mtok":{"value":"not applicable — non-hosted pricing model","citation":null,"status":"not-applicable","tier":"T3"},"cost-output-usd-per-mtok":{"value":"not applicable — non-hosted pricing model","citation":null,"status":"not-applicable","tier":"T3"},"cost-tier":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"cost-pricing-model":{"value":"unknown","citation":null,"status":"estimate","tier":"T3"},"cost-last-verified":{"value":"searched not found","citation":"https://taubench.com/","status":"depth-floor-reached","tier":"T2","source_class":"vendor"},"capability-composite-score":{"value":"(LLM unverified)","citation":null,"status":"estimate","tier":"T3"},"capability-band":{"value":"(LLM unverified)","citation":null,"status":"estimate","tier":"T3"},"capability-benchmark-sources":{"value":"(LLM unverified)","citation":null,"status":"estimate","tier":"T3"},"capability-last-verified":{"value":"(LLM unverified)","citation":null,"status":"estimate","tier":"T3"},"use-case-tags":{"value":"(LLM unverified)","citation":null,"status":"estimate","tier":"T3"},"use-case-anti-tags":{"value":"(LLM unverified)","citation":null,"status":"estimate","tier":"T3"},"adoption-accelerator":{"value":"","citation":null,"status":"no-data","tier":"T3"},"architecture-deployment-locus":{"value":"","citation":null,"status":"no-data","tier":"T3"},"licensing-bucket":{"value":"open-source","citation":"https://taubench.com/","status":"real-data","tier":"T2","source_class":"vendor"},"positioning-platform-vs-product":{"value":"","citation":null,"status":"no-data","tier":"T3"},"taxonomy-three-layer-rollup":{"value":"agent-infra","citation":"https://github.com/MrPeppersDev/agent-infrastructure-landscape/blob/main/docs/schema/section-to-three-layer-rollup.json","status":"real-data","tier":"T1","source_class":"repo"},"developer-activity-opendigger":{"value":"","citation":null,"status":"no-data","tier":"T3"},"freshness-date":{"value":"2026-10-29","citation":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/bench-tau-bench-bench--taubench-com","status":"real-data","tier":"T2","source_class":"catalog"},"compliance-certifications":{"value":"not applicable — no hosted service","citation":"https://github.com/sierra-research/tau2-bench","status":"not-applicable","tier":"T1","source_class":"repo"},"sdk-languages":{"value":"python","citation":"https://raw.githubusercontent.com/sierra-research/tau2-bench/main/README.md","status":"real-data","tier":"T2","source_class":"repo"},"protocol-support":{"value":"","citation":null,"status":"no-data","tier":"T3"},"maturity":{"value":"","citation":null,"status":"no-data","tier":"T3"}},"corroboration":{"cited_cells":12,"third_party_cells":0},"edges":{"outgoing":[],"incoming":[]},"history":[{"date":"2026-10-30","commit":"98ac84a5b9","pr":347,"kind":"added","changes":[]}],"license":"CC-BY-4.0","source":"https://github.com/MrPeppersDev/agent-infrastructure-landscape/blob/main/data/landscape.json"}