{
  "data": {
    "a": {
      "slug": "openai-agents-sdk",
      "name": "OpenAI Agents SDK",
      "vendor": "OpenAI",
      "vendorUrl": "https://openai.com",
      "company": "openai.com",
      "kind": "framework",
      "category": "frameworks",
      "summary": "Multi-agent framework built on agents, hand-offs and guardrails, with sessions, tracing and human approval.",
      "url": "https://www.anchorterminal.com/tools/openai-agents-sdk",
      "markdownUrl": "https://www.anchorterminal.com/tools/openai-agents-sdk.md",
      "slimMarkdownUrl": "https://www.anchorterminal.com/tools/openai-agents-sdk.min.md",
      "jsonUrl": "https://www.anchorterminal.com/api/v1/tools/openai-agents-sdk.json",
      "dns": "openai-agents-sdk.grade.anchorterminal.com",
      "repo": "https://github.com/openai/openai-agents-python",
      "license": "MIT",
      "transports": [],
      "packages": [
        {
          "registry": "pypi",
          "name": "openai-agents"
        },
        {
          "registry": "npm",
          "name": "@openai/agents",
          "main": true
        }
      ],
      "auth": "api-key",
      "authNotes": "OpenAI key by default. Other providers through LiteLLM or any-llm.",
      "pricing": "free",
      "pricingNotes": "Free and open source. You pay for the model calls it makes.",
      "priceSummary": "Free · OSS",
      "where": "library",
      "x402": {
        "level": "no",
        "endpoints": []
      },
      "toolCount": null,
      "popularity": {
        "githubStars": 29709,
        "npmWeekly": 1291899,
        "pypiWeekly": 3018052,
        "asOf": "2026-09-26"
      },
      "docsUrl": "https://openai.github.io/openai-agents-python/",
      "llmsTxt": "https://openai.github.io/openai-agents-python/llms.txt",
      "capabilities": [
        "agent.framework",
        "agent.multi-agent",
        "agent.durable",
        "agent.mcp-client"
      ],
      "tags": [
        "official",
        "framework",
        "python",
        "typescript",
        "open-source",
        "telemetry-default-on"
      ],
      "lastRelease": "2026-10-15",
      "graded": true,
      "anchor": {
        "graded": true,
        "score": 90,
        "grade": "AA",
        "agentReady": true,
        "rank": 3,
        "ranked": true,
        "rankOf": 1557,
        "categoryRank": 2,
        "methodology": "1.0",
        "run": "2026-10-01",
        "scores": {
          "ergonomics": 97,
          "maintenance": 100,
          "payments": 90,
          "reliability": 85,
          "schema": 95,
          "security": 80,
          "transparency": 89
        },
        "assessment": {
          "confidence": "high",
          "date": "2026-11-01"
        },
        "negative": 0,
        "verdict": "MCP in about 11 lines, with static and dynamic tool filters and require_approval. Tracing on by default, with model and tool content, sent to OpenAI.",
        "bestFor": "Teams that want hand-offs, guardrails and approvals with little code, and are happy with OpenAI's tracing or will turn it off.",
        "strengths": [
          "MCP in about 11 lines, with static and dynamic tool filters and require_approval",
          "Input and output guardrails, built-in human approval and sandbox agents",
          "Typed exceptions plus error_handlers, and RunState to resume a paused run",
          "9 open issues and 2 open pull requests on 2026-10-04",
          "A written 0.Y.Z versioning policy with breaking changes listed per minor"
        ],
        "weaknesses": [
          "Tracing on by default, with model and tool content, sent to OpenAI",
          "Tracing isn't available to zero-data-retention organisations",
          "Pre-1.0, and minor releases can carry breaking changes, as 0.20.0, 0.21.0 and 0.22.0 did",
          "0.20.0 changed the default model",
          "Before 0.24.0, RunConfig(tracing_disabled=True) didn't stop spans inside a caller's own trace"
        ],
        "agentNotes": [
          "Set OPENAI_AGENTS_DISABLE_TRACING=1 or OPENAI_AGENTS_TRACE_INCLUDE_SENSITIVE_DATA=0. On versions before 0.24.0, RunConfig(tracing_disabled=True) doesn't stop spans inside your own trace() block",
          "Pin to a minor version. Each 0.Y can break",
          "Set require_approval on MCP servers that write",
          "Name the model explicitly. The default changed in 0.20.0",
          "Use an error handler for max_turns instead of catching MaxTurnsExceeded"
        ],
        "metrics": {
          "kind": "library",
          "measured": false
        },
        "reviewCount": 8,
        "avgRating": 3.9,
        "history": [
          {
            "basis": "public evidence",
            "confidence": "high",
            "grade": "AA",
            "methodology": "1.0",
            "run": "2026-10-01",
            "runLabel": "October 2026 research run",
            "score": 90
          }
        ],
        "editorialScores": {
          "ergonomics": 97,
          "maintenance": 100,
          "payments": 90,
          "reliability": 85,
          "schema": 95,
          "security": 80,
          "transparency": 84
        },
        "provenanceScore": 94,
        "checked": {
          "categories": {
            "ergonomics": "2026-11-01",
            "maintenance": "2026-11-01",
            "payments": "2026-11-01",
            "reliability": "2026-11-01",
            "schema": "2026-11-01",
            "security": "2026-11-01",
            "transparency": "2026-11-01"
          },
          "last": "2026-11-01",
          "lastMoved": "2026-11-13",
          "oldest": "2026-11-01",
          "stale": false,
          "staleAfterDays": 90
        }
      },
      "connect": {
        "install": "pip install openai-agents   # or: npm i @openai/agents"
      },
      "letme": {
        "capability": "https://letme.dev/agent.framework",
        "tool": "https://letme.dev/openai-agents-sdk"
      },
      "sameCompany": [
        "gpt-6-2-astra",
        "openai-api",
        "openai-embeddings",
        "openai-guardrails",
        "openai-moderation",
        "openai-image-api",
        "openai-sora",
        "openai-speech-to-text",
        "gpt-live",
        "openai-realtime",
        "openai-decisions-api",
        "openai-agents-api",
        "openai-codex",
        "openai-dots",
        "gpt-oss-120b",
        "gpt-oss-20b"
      ],
      "area": "frameworks",
      "provenance": {
        "legalEntity": "OpenAI OpCo, LLC",
        "domain": "openai.com",
        "domainRegistered": "2007-01-19",
        "endpointOnVendorDomain": null,
        "terms": "https://openai.com/policies/services-agreement/",
        "privacy": "https://openai.com/policies/privacy-policy/",
        "statusPage": "https://status.openai.com",
        "changelog": "https://openai.github.io/openai-agents-python/release/",
        "securityTxt": "valid",
        "checked": "2026-09-26",
        "score": 94
      },
      "requirements": {
        "setupChecked": true,
        "complianceChecked": true,
        "card": "unknown",
        "freeTier": true,
        "minutes": 5,
        "credential": "api-key",
        "hosting": [
          "self-hosted"
        ],
        "euResidency": "unknown",
        "trainsOnData": "no",
        "zeroRetention": "unknown",
        "dpa": true,
        "selfHost": "yes"
      },
      "pageJsonUrl": "https://www.anchorterminal.com/tools/openai-agents-sdk.json",
      "live": {
        "slug": "openai-agents-sdk",
        "vendorStatus": {
          "page": "https://status.openai.com",
          "indicator": "none",
          "summary": "All Systems Operational",
          "checkedAt": "2026-11-18T23:00:52.098185472Z"
        },
        "versions": [
          {
            "registry": "github",
            "name": "openai/openai-agents-python",
            "version": "v0.25.0",
            "released": "2026-11-18",
            "seenAt": "2026-11-18T18:14:04.517535837Z"
          },
          {
            "registry": "npm",
            "name": "@openai/agents",
            "version": "0.22.0",
            "seenAt": "2026-11-18T18:14:00.253776553Z"
          },
          {
            "registry": "pypi",
            "name": "openai-agents",
            "version": "0.25.0",
            "released": "2026-11-18",
            "seenAt": "2026-11-18T18:13:55.806535668Z"
          }
        ],
        "githubStars": 30753,
        "npmWeekly": 2413850,
        "pypiWeekly": 4755520,
        "securityTxt": {
          "url": "https://openai.com/.well-known/security.txt",
          "state": "valid",
          "checkedAt": "2026-11-18T16:29:37.978845519Z"
        },
        "llmsTxt": {
          "url": "https://openai.github.io/openai-agents-python/llms.txt",
          "ok": true,
          "status": 200,
          "checkedAt": "2026-11-18T16:15:29.221972795Z"
        },
        "domain": {
          "domain": "openai.com",
          "registered": "2007-01-19",
          "source": "https://rdap.verisign.com/com/v1/domain/openai.com",
          "checkedAt": "2026-11-15T15:17:47.16364371Z"
        },
        "pages": [
          {
            "url": "https://openai.github.io/openai-agents-python/release/",
            "kind": "changelog",
            "status": 200,
            "checkedAt": "2026-11-18T21:49:34.85735776Z",
            "changedAt": "2026-10-28T04:39:34.685607508Z",
            "fingerprint": "251fd383498a"
          },
          {
            "url": "https://openai.com/policies/privacy-policy/",
            "kind": "privacy",
            "status": 403,
            "checkedAt": "2026-10-16T22:27:51.367579814Z",
            "changedAt": "2026-10-11T19:31:08.710809907Z",
            "fingerprint": "604e9f69c16a"
          },
          {
            "url": "https://openai.com/policies/services-agreement/",
            "kind": "terms",
            "status": 403,
            "checkedAt": "2026-10-16T22:27:53.401505958Z",
            "changedAt": "2026-10-04T15:46:24.458479472Z",
            "fingerprint": "46a4b30744ca"
          }
        ],
        "github": {
          "repo": "openai/openai-agents-python",
          "checkedAt": "2026-11-16T12:39:09.041917912Z",
          "stars": 30712,
          "archived": false,
          "issuesEnabled": true,
          "pushedAt": "2026-11-16T05:13:15Z",
          "openIssues": 32,
          "openPRs": 14,
          "issuesOpened90d": 261,
          "issuesClosed90d": 238,
          "prsMerged90d": 880,
          "ci": {
            "branch": "main",
            "runs": 50,
            "success": 48,
            "failure": 0,
            "other": 2,
            "latest": "success",
            "latestAt": "2026-11-11T13:10:10Z"
          },
          "replies": {
            "sampled": 25,
            "unanswered": 11,
            "medianDays": 1.4
          }
        },
        "updatedAt": "2026-11-18T23:00:52.098185472Z"
      }
    },
    "answer": "OpenAI Agents SDK scores 90 (AA) on agent readiness against Terse Durable Actors's 69.6 (B), and leads in 6 of 7 scored categories.",
    "b": {
      "slug": "terse-durable-actors",
      "name": "Terse Durable Actors",
      "vendor": "Terse",
      "vendorUrl": "https://useterse.ai/",
      "company": "useterse.ai",
      "kind": "framework",
      "category": "frameworks",
      "summary": "Durable Actors is Terse's MIT-licensed runtime for stateful serverless actors with durable state and one-call-at-a-time execution, with TypeScript and Python SDKs, an HTTP API and an MCP endpoint on the control plane you run.",
      "url": "https://www.anchorterminal.com/tools/terse-durable-actors",
      "markdownUrl": "https://www.anchorterminal.com/tools/terse-durable-actors.md",
      "slimMarkdownUrl": "https://www.anchorterminal.com/tools/terse-durable-actors.min.md",
      "jsonUrl": "https://www.anchorterminal.com/api/v1/tools/terse-durable-actors.json",
      "dns": "terse-durable-actors.grade.anchorterminal.com",
      "repo": "https://github.com/TerseAI/durable-actors",
      "license": "MIT",
      "transports": [
        "streamable-http"
      ],
      "packages": [
        {
          "registry": "npm",
          "name": "durable-actors"
        },
        {
          "registry": "pypi",
          "name": "durable-actors"
        }
      ],
      "auth": "api-key",
      "authNotes": "Self-hosted. The server secret `DURABLE_ACTORS_SECRET` is sent as a bearer token to the HTTP API and the MCP endpoint and grants everything. An optional `DURABLE_ACTORS_INVOKE_API_KEY` only finds and invokes public actor methods. With no secret set, the API is unauthenticated. Local development needs no credential. Terse Cloud sign-in was not read.",
      "pricing": "freemium",
      "pricingNotes": "The runtime and SDKs are free under MIT and run on your own Google Cloud project with no account or card. Terse Cloud bills actors by the second for the CPU and memory they request, US$0.00001971 per CPU-second and US$0.00000667 per GiB-second, per the pricing page. No free allowance or card terms for Terse Cloud were found.",
      "priceSummary": "$0.071 / vCPU-hr",
      "where": "library",
      "x402": {
        "level": "no",
        "evidence": "No x402, MPP or L402 in the repository's code or docs (checked 2026-11-18). Terse Cloud was not read.",
        "endpoints": []
      },
      "toolCount": 10,
      "popularity": {
        "githubStars": 368,
        "npmWeekly": null,
        "pypiWeekly": null,
        "asOf": "2026-11-18"
      },
      "docsUrl": "https://docs.useterse.ai",
      "llmsTxt": "https://docs.useterse.ai/llms.txt",
      "openapi": "https://github.com/TerseAI/durable-actors/blob/main/docs/reference/openapi.yaml",
      "capabilities": [
        "agent.durable",
        "agent.framework",
        "agent.multi-agent"
      ],
      "tags": [
        "framework",
        "open-source",
        "self-hosted",
        "mcp",
        "openapi",
        "typescript",
        "python",
        "no-card",
        "pre-1.0",
        "freemium",
        "usage"
      ],
      "lastRelease": "2026-11-18",
      "graded": true,
      "anchor": {
        "graded": true,
        "score": 69.6,
        "grade": "B",
        "agentReady": false,
        "rank": 331,
        "ranked": true,
        "rankOf": 1557,
        "categoryRank": 14,
        "methodology": "1.0",
        "run": "2026-10-01",
        "scores": {
          "ergonomics": 89,
          "maintenance": 83,
          "payments": 90,
          "reliability": 51,
          "schema": 81,
          "security": 49,
          "transparency": 53
        },
        "assessment": {
          "confidence": "medium",
          "date": "2026-11-18"
        },
        "negative": 0,
        "warnings": [
          {
            "kind": "operator-unidentified",
            "name": "Operator not identified",
            "text": "Terse Cloud's terms and privacy policy name Terse Inc., under Delaware law, and the site footer says New York, with no state of registration, registration number or address. OpenCorporates refused our reader (403), SEC EDGAR's search is closed to it by robots.txt, and Delaware's entity search needs a form it can't submit, so no register entry was found.",
            "deduct": 0,
            "capGrade": "B",
            "since": "2026-11-18",
            "sources": [
              "https://www.useterse.ai/terms",
              "https://www.useterse.ai/privacy",
              "https://www.useterse.ai/pricing",
              "https://www.useterse.ai/"
            ]
          }
        ],
        "verdict": "Durable Actors gives coding agents ten compact MCP tools with read-only and destructive hints, a public OpenAPI 3.1 file and typed TypeScript and Python clients, under MIT with near-daily releases. It is pre-1.0 at 0.18.7, one server secret controls everything unless an invoke-only key is added, and Terse's policies predate Terse Cloud.",
        "bestFor": "Stateful actors for chat, collaboration and multi-agent coordination, as an open-source alternative to Cloudflare Durable Objects per its README.",
        "strengths": [
          "Ten MCP tools on the control plane, nine marked `readOnlyHint` and `invoke_actor` marked `destructiveHint`, with typed, paged inputs",
          "OpenAPI 3.1 file in the repository at docs/reference/openapi.yaml, with an error object of code and message",
          "MIT licence, and 118 tags from v0.1.0 on 2 September 2026 to v0.18.7 on 18 November 2026",
          "SDK telemetry is off by default and goes to standard error, and trace export happens only when an endpoint is configured",
          "Start limits answer 429 with `actor_quota_exceeded` or `actor_start_rate_limited`, the latter with `Retry-After`"
        ],
        "weaknesses": [
          "Version 0.18.7, with no stable release and no changelog file in the repository",
          "A server with `DURABLE_ACTORS_SECRET` unset accepts unauthenticated API requests, per SECURITY.md",
          "One server secret grants deploy, observe and invoke. The only narrower credential is an optional invoke-only key",
          "Terse's terms and privacy policy (14 May 2026) describe an earlier product and don't cover Terse Cloud",
          "Terse Inc. names no state of registration or address, and no register entry was found"
        ],
        "agentNotes": [
          "Call `list_projects` first and pass its `projectId` to every other tool",
          "Never retry `invoke_actor` when the outcome is unknown, because calls are not deduplicated and run real code",
          "Set `DURABLE_ACTORS_SECRET` before exposing a server beyond localhost, and send it as a bearer token",
          "Give backends that only call actors the invoke-only key, not the server secret",
          "On 429 `actor_start_rate_limited`, wait for the `Retry-After` seconds before starting another actor"
        ],
        "metrics": {
          "kind": "local",
          "measured": false
        },
        "reviewCount": 0,
        "avgRating": 0,
        "history": [
          {
            "basis": "public evidence",
            "confidence": "medium",
            "grade": "B",
            "methodology": "1.0",
            "run": "2026-10-01",
            "runLabel": "October 2026 research run",
            "score": 69.6
          }
        ],
        "editorialScores": {
          "ergonomics": 89,
          "maintenance": 83,
          "payments": 90,
          "reliability": 51,
          "schema": 81,
          "security": 49,
          "transparency": 58
        },
        "provenanceScore": 48,
        "checked": {
          "categories": {
            "ergonomics": "2026-11-18",
            "maintenance": "2026-11-18",
            "payments": "2026-11-18",
            "reliability": "2026-11-18",
            "schema": "2026-11-18",
            "security": "2026-11-18",
            "transparency": "2026-11-18"
          },
          "last": "2026-11-18",
          "lastMoved": "2026-11-18",
          "oldest": "2026-11-18",
          "stale": false,
          "staleAfterDays": 90
        }
      },
      "connect": {
        "install": "bunx durable-actors init my-actors \u0026\u0026 cd my-actors \u0026\u0026 bun install \u0026\u0026 bunx durable-actors dev",
        "claudeCode": "claude mcp add --transport http durable-actors http://127.0.0.1:7100/mcp",
        "config": {
          "mcpServers": {
            "durable-actors": {
              "url": "http://127.0.0.1:7100/mcp"
            }
          }
        }
      },
      "letme": {
        "capability": "https://letme.dev/agent.durable",
        "tool": "https://letme.dev/terse-durable-actors"
      },
      "area": "frameworks",
      "unitPrices": [
        {
          "item": "Terse Cloud actor CPU requested",
          "unit": "vcpu-hour",
          "usd": 0.070956,
          "note": "Published as US$0.00001971 per CPU per second, times 3,600. Memory is billed on top at US$0.00000667 per GiB per second"
        }
      ],
      "provenance": {
        "legalEntity": "Terse Inc.",
        "domain": "useterse.ai",
        "domainRegistered": "",
        "endpointOnVendorDomain": false,
        "terms": "https://www.useterse.ai/terms",
        "privacy": "https://www.useterse.ai/privacy",
        "statusPage": "",
        "changelog": "https://github.com/TerseAI/durable-actors/releases",
        "securityTxt": "unknown",
        "checked": "2026-11-18",
        "notes": [
          "The terms and privacy policy, both dated 14 May 2026, name Terse Inc. (doing business as Terse), the terms under Delaware law. The site footer reads © 2026 Terse · New York. No state of registration, registration number or address was found.",
          "Both documents describe an earlier Terse product of flows, workspaces and credits, and neither mentions Durable Actors or Terse Cloud. The terms forbid using any robot, spider or other automatic means to access the Service, recorded here with no deduction.",
          "useterse.ai's robots.txt allows every agent except /thank-you and /mellon. docs.useterse.ai's allows all agents with Content-Signal ai-input=yes and serves llms.txt (read 18 November 2026).",
          "The repository was read with git clone. SECURITY.md gives security@useterse.ai for private reports."
        ],
        "score": 48
      },
      "pageJsonUrl": "https://www.anchorterminal.com/tools/terse-durable-actors.json"
    },
    "edges": [
      {
        "by": 36,
        "category": "transparency",
        "leads": "openai-agents-sdk",
        "name": "Transparency \u0026 trust",
        "reasons": {
          "a": "MIT (30). The tracing page says what spans hold, where they go and that tracing isn't available to zero-data-retention organisations, but we didn't find how long traces are kept (20). Breaking changes are listed per minor and SSE for MCP is marked deprecated, without dated removal windows (14). Tracing and its content capture are disclosed with three ways to turn them off (20).",
          "b": "MIT, copyright 2026 Terse (30). Self-hosted state stays in your PostgreSQL database and GCS bucket per the self-hosting guide. Terse's terms and privacy policy, both dated 14 May 2026, describe an earlier product of flows and workspaces, give no retention periods and don't mention Terse Cloud (10 of 30). No deprecation policy found, and SECURITY.md says older releases get no backports (0 of 20). SDK invocation telemetry is off by default and writes to standard error when enabled, and request spans and platform events are exported only to a collector the operator configures. No call to Terse's own hosts was found in the runtime or SDK source (18 of 20). Total 58."
        },
        "scores": {
          "a": 89,
          "b": 53
        }
      },
      {
        "by": 34,
        "category": "reliability",
        "leads": "openai-agents-sdk",
        "name": "Reliability",
        "reasons": {
          "a": "Official packages on PyPI (Python 3.10 or newer, `requires-python` re-read on 1 November) and npm (20). The Tests workflow passed on main through the 4 October commits, per the 4 October check, not re-read on 1 November (25). 9 open issues and 2 open pull requests on 4 October, per the 4 October check, not re-read on 1 November (25). A written policy for 0.Y.Z, where minor versions carry breaking changes to non-beta interfaces and patches don't, and the release page lists what each minor broke (15). 0.23.0 and 0.24.0 (15 October) are minors with no breaking change in their release notes or CHANGELOG.md, and the release page still has no entry for either on 1 November. 0.24.0, still pre-1.0 and still the newest on PyPI on 1 November (0).",
          "b": "Graded as local software, the runtime and SDKs you run yourself. The README and the Python SDK README install with `bunx durable-actors` or `uvx --from 'durable-actors[cli]'` and state Bun 1.4.2+ and Python 3.11+, and the release workflow publishes to npm. The registry pages were not read (15 of 20). A public CI workflow runs Rust, Python, TypeScript and example builds on every push and pull request. Whether it passes on main was not checked (15 of 25). 23 issues were open on 18 November 2026, mostly improvements and feature requests, with a few bug reports such as a constructor that keeps its SQLite writes after throwing (17 November) and alarms that run more than once (11 November) (18 of 25). Release notes are generated from pull-request labels on GitHub Releases, which were not read, and no changelog file or breaking-change notes were found in the repository (3 of 15). Version 0.18.7, not 1.0 or declared stable (0). Total 51."
        },
        "scores": {
          "a": 85,
          "b": 51
        }
      },
      {
        "by": 31,
        "category": "security",
        "leads": "openai-agents-sdk",
        "name": "Security \u0026 auth",
        "reasons": {
          "a": "Tracing is on by default and trace_include_sensitive_data defaults to true, so model and function-call inputs and outputs go to OpenAI's Traces dashboard. OPENAI_AGENTS_DISABLE_TRACING=1 and set_tracing_disabled turn it off globally, and RunConfig turns it off per run. Through 0.23.1 the per-run opt-out didn't apply inside the caller's own trace. The fix shipped in 0.24.0 on 15 October (10). Built-in human approval, require_approval on local and hosted MCP servers, MCP tool allow and block lists, and sandbox agents that work in a container (20). Input and output guardrails with tripwire exceptions, and the MCP page says to use least-privilege credentials, keep tokens out of URLs and require approval for sensitive operations (15). Built-in tracing with third-party processors such as Weights \u0026 Biases and Datadog (15). SECURITY.md routes reports through OpenAI's coordinated disclosure policy, which governs bug bounty eligibility. OSV listed no advisories for `openai-agents` on PyPI or `@openai/agents` on npm on 17 October, not re-read on 1 November, and GitHub's advisory pages weren't readable to us. On main, unreleased on 1 November, pyproject.toml raises the `mcp` floor to 1.30.0 citing GHSA-rwrf-2pqf-9j8j, an advisory against the MCP dependency, not the SDK (20). Framework reading, so SOC 2 isn't scored.",
          "b": "One server secret, `DURABLE_ACTORS_SECRET`, as a bearer token for the API and MCP, plus an optional invoke-only key limited to public actor methods. With the secret unset the API is unauthenticated, which SECURITY.md warns about (15 of 30). The invoke-only key and the MCP read-only hints give some least privilege, with no confirmation step before `invoke_actor` (10 of 20). The OpenAPI file marks actor output as untrusted and to be rendered as plain text, with no guidance for agents on injected instructions (6 of 15). Request history, actor logs, WebSocket sessions and state versions tied to the request that saved them (12 of 15). SECURITY.md gives a private reporting address, security@useterse.ai. No bug bounty, certification or published advisories found (6 of 20). On GKE the self-hosting guide runs actors in gVisor sandbox pods with per-actor storage credentials. Total 49."
        },
        "scores": {
          "a": 80,
          "b": 49
        }
      },
      {
        "by": 17,
        "category": "maintenance",
        "leads": "openai-agents-sdk",
        "name": "Maintenance \u0026 community",
        "reasons": {
          "a": "0.24.0 on PyPI on 2026-10-15, 17 days before 1 November (30). 11 PyPI releases in the 90 days to 1 November, from 0.19.3 on 4 August to 0.24.0 (20). 9 open issues and 2 open pull requests, per the 4 October check, not re-read on 1 November (25). Python 0.24.0 and @openai/agents 0.21.0 (tagged 20 October) both current (15). Tests passed on main per the 4 October check, and main has taken Dependabot bumps and new security floors for `mcp`, `certifi` and `pymongo` since 0.24.0 (10).",
          "b": "v0.18.7 tagged on 18 November 2026, the day of the check (30). 118 tags since v0.1.0 on 2 September 2026, with eight between 15 and 18 November (20). 23 open issues, three of them with long threads (7, 9 and 15 replies) and most with none. Pull requests from five outside contributors were merged between 2 October and 3 November 2026, and CONTRIBUTING.md says GitHub issues are triaged into the team's tracker (10 of 25). Current official TypeScript and Python SDKs at 0.18.7. The MCP endpoint is self-hosted and not in the official registry, which was not searched (15). CI with pinned action versions and lockfiles for Bun, Cargo and uv (8 of 10). Total 83."
        },
        "scores": {
          "a": 100,
          "b": 83
        }
      },
      {
        "by": 14,
        "category": "schema",
        "leads": "openai-agents-sdk",
        "name": "Schema \u0026 documentation",
        "reasons": {
          "a": "Typed Python API with a reference section in the docs (25). llms.txt with an API Reference Highlights section, re-read on 1 November (10). Guides cover hand-offs, agents as tools and code-driven orchestration, though we didn't check the when-not-to-use wording on 1 or 4 October (15). Function tools get their schemas from typed Python signatures (15). Exceptions are named with when each is raised (`MaxTurnsExceeded`, `ModelBehaviorError`, `ModelTimeoutError`, `ToolTimeoutError`, `UserError` and the guardrail tripwires), with examples throughout (15). Versioning policy and release notes per minor, and from 0.23.0 a release-please CHANGELOG.md in the repository (15).",
          "b": "OpenAPI 3.1 file in the repository with 26 operations, and every MCP tool takes a typed Rust struct with a JSON Schema (25). Markdown docs in the repository, an AGENTS.md and an agent skill for setting up projects (10). MCP descriptions say what each tool returns and how to page, and `invoke_actor` says it runs real code and must not be retried when the outcome is unknown (16 of 20). Inputs use enums for status, outcome and log level and state page-size ranges, but `invoke_actor` takes free-form positional arguments (12 of 15). Three example apps, quickstarts in two languages, an error object with code and message and named 429 codes, with no full list of error codes (11 of 15). Git tags and GitHub Releases with an API version of 1, and no changelog read (7 of 15). Total 81."
        },
        "scores": {
          "a": 95,
          "b": 81
        }
      },
      {
        "by": 8,
        "category": "ergonomics",
        "leads": "openai-agents-sdk",
        "name": "Agent ergonomics",
        "reasons": {
          "a": "An agent with one MCP server is about 11 lines, with static and dynamic tool filters and tool-list caching (25). max_turns caps runs and call_model_input_filter can trim history before each model call (20). Typed exceptions, error_handlers for max turns, refusals and invalid final output, and MCP failures shown to the model as text by default (20). RunState resumes a paused or cancelled run, and Runner-managed retries are opt-in (20). An agent needs a name and instructions, but 0.20.0 changed the default model (7). Python and JavaScript (5).",
          "b": "Ten MCP tools with one-line descriptions (25). Cursor paging with stated limits (1 to 100, or 1 to 500 for requests) and filters by actor, outcome, log level, text and time window (20). Errors carry a code and a message, start limits return named codes with `Retry-After`, and invalid MCP input says which characters are allowed (14 of 20). Nine tools carry `readOnlyHint` and `invoke_actor` carries `destructiveHint`, but invocations have no idempotency key (15 of 20). Local development needs one command and no credential, with official TypeScript and Python SDKs and generated typed clients (15). Total 89."
        },
        "scores": {
          "a": 97,
          "b": 89
        }
      }
    ],
    "facts": [
      {
        "a": "Agent framework",
        "b": "Agent framework",
        "name": "Kind"
      },
      {
        "a": "OpenAI",
        "b": "Terse",
        "name": "Vendor"
      },
      {
        "a": "no (local only)",
        "b": "no (local only)",
        "name": "Hosted endpoint"
      },
      {
        "a": "",
        "b": "Streamable HTTP",
        "name": "Transports"
      },
      {
        "a": "API key",
        "b": "API key",
        "name": "Auth"
      },
      {
        "a": "Free",
        "b": "Freemium",
        "name": "Pricing"
      },
      {
        "a": "no",
        "b": "no",
        "name": "x402"
      },
      {
        "a": "MIT",
        "b": "MIT",
        "name": "Licence"
      },
      {
        "a": "none",
        "b": "10",
        "name": "Tools exposed"
      },
      {
        "a": "no",
        "b": "no",
        "name": "Read-only variant documented"
      },
      {
        "a": "yes",
        "b": "yes",
        "name": "llms.txt"
      },
      {
        "a": "2026-10-15",
        "b": "2026-11-18",
        "name": "Last release"
      },
      {
        "a": "couldn't be read",
        "b": "2026-05-14",
        "name": "Terms last updated"
      },
      {
        "a": "couldn't be read",
        "b": "2026-05-14",
        "name": "Privacy policy last updated"
      },
      {
        "a": "couldn't be read",
        "b": "not found in the text",
        "name": "Customer content may train models"
      },
      {
        "a": "couldn't be read",
        "b": "yes",
        "name": "Terms restrict automated access"
      },
      {
        "a": "couldn't be read",
        "b": "not found in the text",
        "name": "Terms restrict benchmarking"
      },
      {
        "a": "couldn't be read",
        "b": "yes",
        "name": "Terms or service can change without notice"
      },
      {
        "a": "couldn't be read",
        "b": "not found in the text",
        "name": "Arbitration or class-action waiver"
      },
      {
        "a": "30k stars, 1.3M npm/wk, 3M PyPI/wk",
        "b": "368 stars",
        "name": "Popularity"
      },
      {
        "a": "3.9/5 (8)",
        "b": "none",
        "name": "Agent reviews"
      }
    ],
    "faq": [
      {
        "answer": "OpenAI Agents SDK scores 90 (AA) on agent readiness against Terse Durable Actors's 69.6 (B), and leads in 6 of 7 scored categories.",
        "question": "Which is better for AI agents, OpenAI Agents SDK or Terse Durable Actors?"
      },
      {
        "answer": "Yes. OpenAI Agents SDK is open source (MIT). Terse Durable Actors is open source (MIT).",
        "question": "Are OpenAI Agents SDK and Terse Durable Actors open source?"
      }
    ],
    "goodFor": [
      {
        "aheadOn": [
          "Reliability, 85 against 51",
          "Schema \u0026 documentation, 95 against 81",
          "Agent ergonomics, 97 against 89",
          "Security \u0026 auth, 80 against 49",
          "Maintenance \u0026 community, 100 against 83",
          "Transparency \u0026 trust, 89 against 53"
        ],
        "also": [
          "Agent-ready, a grade of BB or better"
        ],
        "goodFor": "Teams that want hand-offs, guardrails and approvals with little code, and are happy with OpenAI's tracing or will turn it off.",
        "slug": "openai-agents-sdk",
        "watchFor": "Tracing on by default, with model and tool content, sent to OpenAI"
      },
      {
        "aheadOn": null,
        "also": [
          "Free to start without a card"
        ],
        "goodFor": "Stateful actors for chat, collaboration and multi-agent coordination, as an open-source alternative to Cloudflare Durable Objects per its README.",
        "slug": "terse-durable-actors",
        "watchFor": "Version 0.18.7, with no stable release and no changelog file in the repository"
      }
    ],
    "history": {
      "openai-agents-sdk": [
        {
          "causes": [
            "method"
          ],
          "date": "2026-11-13",
          "delta": 3.8,
          "from": 86.2,
          "fromGrade": "AA",
          "to": 90,
          "toGrade": "AA"
        },
        {
          "causes": [
            "method"
          ],
          "date": "2026-10-08",
          "delta": -0.3,
          "from": 86.5,
          "fromGrade": "AA",
          "to": 86.2,
          "toGrade": "AA"
        }
      ],
      "terse-durable-actors": [
        {
          "causes": [
            "terms",
            "pricing",
            "evidence"
          ],
          "date": "2026-11-18",
          "delta": 8.9,
          "from": 60.7,
          "fromGrade": "C",
          "to": 69.6,
          "toGrade": "B"
        }
      ]
    },
    "job": {
      "capability": "agent.framework",
      "name": "Agent frameworks"
    },
    "others": [
      {
        "json": "https://www.anchorterminal.com/compare/agentos-vs-openai-agents-sdk.json",
        "title": "AgentOS vs OpenAI Agents SDK",
        "url": "https://www.anchorterminal.com/compare/agentos-vs-openai-agents-sdk"
      },
      {
        "json": "https://www.anchorterminal.com/compare/agentos-vs-terse-durable-actors.json",
        "title": "AgentOS vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/agentos-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/agno-vs-openai-agents-sdk.json",
        "title": "Agno vs OpenAI Agents SDK",
        "url": "https://www.anchorterminal.com/compare/agno-vs-openai-agents-sdk"
      },
      {
        "json": "https://www.anchorterminal.com/compare/agno-vs-terse-durable-actors.json",
        "title": "Agno vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/agno-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/boxlang-ai-vs-openai-agents-sdk.json",
        "title": "BoxLang AI vs OpenAI Agents SDK",
        "url": "https://www.anchorterminal.com/compare/boxlang-ai-vs-openai-agents-sdk"
      },
      {
        "json": "https://www.anchorterminal.com/compare/boxlang-ai-vs-terse-durable-actors.json",
        "title": "BoxLang AI vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/boxlang-ai-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/claude-agent-sdk-vs-openai-agents-sdk.json",
        "title": "Claude Agent SDK vs OpenAI Agents SDK",
        "url": "https://www.anchorterminal.com/compare/claude-agent-sdk-vs-openai-agents-sdk"
      },
      {
        "json": "https://www.anchorterminal.com/compare/claude-agent-sdk-vs-terse-durable-actors.json",
        "title": "Claude Agent SDK vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/claude-agent-sdk-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/crewai-vs-openai-agents-sdk.json",
        "title": "CrewAI vs OpenAI Agents SDK",
        "url": "https://www.anchorterminal.com/compare/crewai-vs-openai-agents-sdk"
      },
      {
        "json": "https://www.anchorterminal.com/compare/crewai-vs-terse-durable-actors.json",
        "title": "CrewAI vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/crewai-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/cursor-agent-sdk-vs-openai-agents-sdk.json",
        "title": "Cursor Agent SDK vs OpenAI Agents SDK",
        "url": "https://www.anchorterminal.com/compare/cursor-agent-sdk-vs-openai-agents-sdk"
      },
      {
        "json": "https://www.anchorterminal.com/compare/cursor-agent-sdk-vs-terse-durable-actors.json",
        "title": "Cursor Agent SDK vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/cursor-agent-sdk-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/docker-agent-vs-openai-agents-sdk.json",
        "title": "Docker Agent vs OpenAI Agents SDK",
        "url": "https://www.anchorterminal.com/compare/docker-agent-vs-openai-agents-sdk"
      },
      {
        "json": "https://www.anchorterminal.com/compare/docker-agent-vs-terse-durable-actors.json",
        "title": "Docker Agent vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/docker-agent-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/google-adk-vs-openai-agents-sdk.json",
        "title": "Agent Development Kit (ADK) vs OpenAI Agents SDK",
        "url": "https://www.anchorterminal.com/compare/google-adk-vs-openai-agents-sdk"
      },
      {
        "json": "https://www.anchorterminal.com/compare/google-adk-vs-terse-durable-actors.json",
        "title": "Agent Development Kit (ADK) vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/google-adk-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/guilds-run-vs-openai-agents-sdk.json",
        "title": "guilds.run vs OpenAI Agents SDK",
        "url": "https://www.anchorterminal.com/compare/guilds-run-vs-openai-agents-sdk"
      },
      {
        "json": "https://www.anchorterminal.com/compare/guilds-run-vs-terse-durable-actors.json",
        "title": "guilds.run vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/guilds-run-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/langgraph-vs-openai-agents-sdk.json",
        "title": "LangGraph vs OpenAI Agents SDK",
        "url": "https://www.anchorterminal.com/compare/langgraph-vs-openai-agents-sdk"
      },
      {
        "json": "https://www.anchorterminal.com/compare/langgraph-vs-terse-durable-actors.json",
        "title": "LangGraph vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/langgraph-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/llamaindex-vs-openai-agents-sdk.json",
        "title": "LlamaIndex Framework vs OpenAI Agents SDK",
        "url": "https://www.anchorterminal.com/compare/llamaindex-vs-openai-agents-sdk"
      },
      {
        "json": "https://www.anchorterminal.com/compare/llamaindex-vs-terse-durable-actors.json",
        "title": "LlamaIndex Framework vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/llamaindex-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/mastra-vs-openai-agents-sdk.json",
        "title": "Mastra vs OpenAI Agents SDK",
        "url": "https://www.anchorterminal.com/compare/mastra-vs-openai-agents-sdk"
      },
      {
        "json": "https://www.anchorterminal.com/compare/mastra-vs-terse-durable-actors.json",
        "title": "Mastra vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/mastra-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/microsoft-agent-framework-vs-openai-agents-sdk.json",
        "title": "Microsoft Agent Framework vs OpenAI Agents SDK",
        "url": "https://www.anchorterminal.com/compare/microsoft-agent-framework-vs-openai-agents-sdk"
      },
      {
        "json": "https://www.anchorterminal.com/compare/microsoft-agent-framework-vs-terse-durable-actors.json",
        "title": "Microsoft Agent Framework vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/microsoft-agent-framework-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/openai-agents-sdk-vs-pydantic-ai.json",
        "title": "OpenAI Agents SDK vs Pydantic AI",
        "url": "https://www.anchorterminal.com/compare/openai-agents-sdk-vs-pydantic-ai"
      },
      {
        "json": "https://www.anchorterminal.com/compare/openai-agents-sdk-vs-smolagents.json",
        "title": "OpenAI Agents SDK vs smolagents",
        "url": "https://www.anchorterminal.com/compare/openai-agents-sdk-vs-smolagents"
      },
      {
        "json": "https://www.anchorterminal.com/compare/openai-agents-sdk-vs-strands-agents.json",
        "title": "OpenAI Agents SDK vs Strands Agents",
        "url": "https://www.anchorterminal.com/compare/openai-agents-sdk-vs-strands-agents"
      },
      {
        "json": "https://www.anchorterminal.com/compare/openai-agents-sdk-vs-tradingagents.json",
        "title": "OpenAI Agents SDK vs TradingAgents",
        "url": "https://www.anchorterminal.com/compare/openai-agents-sdk-vs-tradingagents"
      },
      {
        "json": "https://www.anchorterminal.com/compare/openai-agents-sdk-vs-vercel-ai-sdk.json",
        "title": "OpenAI Agents SDK vs AI SDK",
        "url": "https://www.anchorterminal.com/compare/openai-agents-sdk-vs-vercel-ai-sdk"
      },
      {
        "json": "https://www.anchorterminal.com/compare/pydantic-ai-vs-terse-durable-actors.json",
        "title": "Pydantic AI vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/pydantic-ai-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/smolagents-vs-terse-durable-actors.json",
        "title": "smolagents vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/smolagents-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/strands-agents-vs-terse-durable-actors.json",
        "title": "Strands Agents vs Terse Durable Actors",
        "url": "https://www.anchorterminal.com/compare/strands-agents-vs-terse-durable-actors"
      },
      {
        "json": "https://www.anchorterminal.com/compare/terse-durable-actors-vs-tradingagents.json",
        "title": "Terse Durable Actors vs TradingAgents",
        "url": "https://www.anchorterminal.com/compare/terse-durable-actors-vs-tradingagents"
      },
      {
        "json": "https://www.anchorterminal.com/compare/terse-durable-actors-vs-vercel-ai-sdk.json",
        "title": "Terse Durable Actors vs AI SDK",
        "url": "https://www.anchorterminal.com/compare/terse-durable-actors-vs-vercel-ai-sdk"
      }
    ],
    "panel": null,
    "scores": [
      {
        "by": 34,
        "edge": "openai-agents-sdk",
        "key": "reliability",
        "name": "Reliability",
        "openai-agents-sdk": 85,
        "terse-durable-actors": 51,
        "weight": 16
      },
      {
        "key": "performance",
        "name": "Performance",
        "pending": true,
        "weight": 10
      },
      {
        "by": 14,
        "edge": "openai-agents-sdk",
        "key": "schema",
        "name": "Schema \u0026 documentation",
        "openai-agents-sdk": 95,
        "terse-durable-actors": 81,
        "weight": 13
      },
      {
        "by": 8,
        "edge": "openai-agents-sdk",
        "key": "ergonomics",
        "name": "Agent ergonomics",
        "openai-agents-sdk": 97,
        "terse-durable-actors": 89,
        "weight": 13
      },
      {
        "by": 31,
        "edge": "openai-agents-sdk",
        "key": "security",
        "name": "Security \u0026 auth",
        "openai-agents-sdk": 80,
        "terse-durable-actors": 49,
        "weight": 14
      },
      {
        "by": 0,
        "edge": "",
        "key": "payments",
        "name": "Payments \u0026 pricing",
        "openai-agents-sdk": 90,
        "terse-durable-actors": 90,
        "weight": 10
      },
      {
        "key": "tasks",
        "name": "Task success",
        "pending": true,
        "weight": 10
      },
      {
        "by": 17,
        "edge": "openai-agents-sdk",
        "key": "maintenance",
        "name": "Maintenance \u0026 community",
        "openai-agents-sdk": 100,
        "terse-durable-actors": 83,
        "weight": 7
      },
      {
        "by": 36,
        "edge": "openai-agents-sdk",
        "key": "transparency",
        "name": "Transparency \u0026 trust",
        "openai-agents-sdk": 89,
        "terse-durable-actors": 53,
        "weight": 7
      }
    ],
    "summary": "OpenAI Agents SDK scores 90 (AA) on agent readiness against Terse Durable Actors's 69.6 (B), and leads in 6 of 7 scored categories. Both do agent frameworks.",
    "tasks": null,
    "tutorials": [],
    "verdicts": {
      "openai-agents-sdk": "MCP in about 11 lines, with static and dynamic tool filters and require_approval. Tracing on by default, with model and tool content, sent to OpenAI.",
      "terse-durable-actors": "Durable Actors gives coding agents ten compact MCP tools with read-only and destructive hints, a public OpenAPI 3.1 file and typed TypeScript and Python clients, under MIT with near-daily releases. It is pre-1.0 at 0.18.7, one server secret controls everything unless an invoke-only key is added, and Terse's policies predate Terse Cloud."
    }
  },
  "kind": "anchor.page",
  "links": {
    "api": "https://www.anchorterminal.com/api/v1/index.json",
    "html": "https://www.anchorterminal.com/compare/openai-agents-sdk-vs-terse-durable-actors",
    "json": "https://www.anchorterminal.com/compare/openai-agents-sdk-vs-terse-durable-actors.json",
    "llms": "https://www.anchorterminal.com/llms.txt",
    "markdown": "https://www.anchorterminal.com/compare/openai-agents-sdk-vs-terse-durable-actors.md",
    "mcp": "https://www.anchorterminal.com/mcp",
    "slim": "https://www.anchorterminal.com/compare/openai-agents-sdk-vs-terse-durable-actors.min.md"
  },
  "markdown": "OpenAI Agents SDK scores 90 (AA) on agent readiness against Terse Durable Actors's 69.6 (B), and leads in 6 of 7 scored categories. Both do agent frameworks.\n\n- OpenAI Agents SDK: grade AA, 90/100, rank #3 of 1557. Markdown https://www.anchorterminal.com/tools/openai-agents-sdk.md · JSON https://www.anchorterminal.com/api/v1/tools/openai-agents-sdk.json\n- Terse Durable Actors: grade B, 69.6/100, rank #331 of 1557. Markdown https://www.anchorterminal.com/tools/terse-durable-actors.md · JSON https://www.anchorterminal.com/api/v1/tools/terse-durable-actors.json\n- Best agent frameworks and SDKs: https://www.anchorterminal.com/best/frameworks/index.md\n- All 190 frameworks comparisons: https://www.anchorterminal.com/compare/frameworks/index.md\n\n## Which one, for what\n\n### OpenAI Agents SDK (AA)\n\nGood for: Teams that want hand-offs, guardrails and approvals with little code, and are happy with OpenAI's tracing or will turn it off.\n\nAhead on:\n- Reliability, 85 against 51\n- Schema \u0026 documentation, 95 against 81\n- Agent ergonomics, 97 against 89\n- Security \u0026 auth, 80 against 49\n- Maintenance \u0026 community, 100 against 83\n- Transparency \u0026 trust, 89 against 53\n\nAlso in its favour:\n- Agent-ready, a grade of BB or better\n\nWatch for: Tracing on by default, with model and tool content, sent to OpenAI\n\n### Terse Durable Actors (B)\n\nGood for: Stateful actors for chat, collaboration and multi-agent coordination, as an open-source alternative to Cloudflare Durable Objects per its README.\n\nAlso in its favour:\n- Free to start without a card\n\nWatch for: Version 0.18.7, with no stable release and no changelog file in the repository\n\n\n## Why the edge\n\n### Transparency \u0026 trust: OpenAI Agents SDK +36\n\n- OpenAI Agents SDK (89): MIT (30). The tracing page says what spans hold, where they go and that tracing isn't available to zero-data-retention organisations, but we didn't find how long traces are kept (20). Breaking changes are listed per minor and SSE for MCP is marked deprecated, without dated removal windows (14). Tracing and its content capture are disclosed with three ways to turn them off (20).\n- Terse Durable Actors (53): MIT, copyright 2026 Terse (30). Self-hosted state stays in your PostgreSQL database and GCS bucket per the self-hosting guide. Terse's terms and privacy policy, both dated 14 May 2026, describe an earlier product of flows and workspaces, give no retention periods and don't mention Terse Cloud (10 of 30). No deprecation policy found, and SECURITY.md says older releases get no backports (0 of 20). SDK invocation telemetry is off by default and writes to standard error when enabled, and request spans and platform events are exported only to a collector the operator configures. No call to Terse's own hosts was found in the runtime or SDK source (18 of 20). Total 58.\n\n### Reliability: OpenAI Agents SDK +34\n\n- OpenAI Agents SDK (85): Official packages on PyPI (Python 3.10 or newer, `requires-python` re-read on 1 November) and npm (20). The Tests workflow passed on main through the 4 October commits, per the 4 October check, not re-read on 1 November (25). 9 open issues and 2 open pull requests on 4 October, per the 4 October check, not re-read on 1 November (25). A written policy for 0.Y.Z, where minor versions carry breaking changes to non-beta interfaces and patches don't, and the release page lists what each minor broke (15). 0.23.0 and 0.24.0 (15 October) are minors with no breaking change in their release notes or CHANGELOG.md, and the release page still has no entry for either on 1 November. 0.24.0, still pre-1.0 and still the newest on PyPI on 1 November (0).\n- Terse Durable Actors (51): Graded as local software, the runtime and SDKs you run yourself. The README and the Python SDK README install with `bunx durable-actors` or `uvx --from 'durable-actors[cli]'` and state Bun 1.4.2+ and Python 3.11+, and the release workflow publishes to npm. The registry pages were not read (15 of 20). A public CI workflow runs Rust, Python, TypeScript and example builds on every push and pull request. Whether it passes on main was not checked (15 of 25). 23 issues were open on 18 November 2026, mostly improvements and feature requests, with a few bug reports such as a constructor that keeps its SQLite writes after throwing (17 November) and alarms that run more than once (11 November) (18 of 25). Release notes are generated from pull-request labels on GitHub Releases, which were not read, and no changelog file or breaking-change notes were found in the repository (3 of 15). Version 0.18.7, not 1.0 or declared stable (0). Total 51.\n\n### Security \u0026 auth: OpenAI Agents SDK +31\n\n- OpenAI Agents SDK (80): Tracing is on by default and trace_include_sensitive_data defaults to true, so model and function-call inputs and outputs go to OpenAI's Traces dashboard. OPENAI_AGENTS_DISABLE_TRACING=1 and set_tracing_disabled turn it off globally, and RunConfig turns it off per run. Through 0.23.1 the per-run opt-out didn't apply inside the caller's own trace. The fix shipped in 0.24.0 on 15 October (10). Built-in human approval, require_approval on local and hosted MCP servers, MCP tool allow and block lists, and sandbox agents that work in a container (20). Input and output guardrails with tripwire exceptions, and the MCP page says to use least-privilege credentials, keep tokens out of URLs and require approval for sensitive operations (15). Built-in tracing with third-party processors such as Weights \u0026 Biases and Datadog (15). SECURITY.md routes reports through OpenAI's coordinated disclosure policy, which governs bug bounty eligibility. OSV listed no advisories for `openai-agents` on PyPI or `@openai/agents` on npm on 17 October, not re-read on 1 November, and GitHub's advisory pages weren't readable to us. On main, unreleased on 1 November, pyproject.toml raises the `mcp` floor to 1.30.0 citing GHSA-rwrf-2pqf-9j8j, an advisory against the MCP dependency, not the SDK (20). Framework reading, so SOC 2 isn't scored.\n- Terse Durable Actors (49): One server secret, `DURABLE_ACTORS_SECRET`, as a bearer token for the API and MCP, plus an optional invoke-only key limited to public actor methods. With the secret unset the API is unauthenticated, which SECURITY.md warns about (15 of 30). The invoke-only key and the MCP read-only hints give some least privilege, with no confirmation step before `invoke_actor` (10 of 20). The OpenAPI file marks actor output as untrusted and to be rendered as plain text, with no guidance for agents on injected instructions (6 of 15). Request history, actor logs, WebSocket sessions and state versions tied to the request that saved them (12 of 15). SECURITY.md gives a private reporting address, security@useterse.ai. No bug bounty, certification or published advisories found (6 of 20). On GKE the self-hosting guide runs actors in gVisor sandbox pods with per-actor storage credentials. Total 49.\n\n### Maintenance \u0026 community: OpenAI Agents SDK +17\n\n- OpenAI Agents SDK (100): 0.24.0 on PyPI on 2026-10-15, 17 days before 1 November (30). 11 PyPI releases in the 90 days to 1 November, from 0.19.3 on 4 August to 0.24.0 (20). 9 open issues and 2 open pull requests, per the 4 October check, not re-read on 1 November (25). Python 0.24.0 and @openai/agents 0.21.0 (tagged 20 October) both current (15). Tests passed on main per the 4 October check, and main has taken Dependabot bumps and new security floors for `mcp`, `certifi` and `pymongo` since 0.24.0 (10).\n- Terse Durable Actors (83): v0.18.7 tagged on 18 November 2026, the day of the check (30). 118 tags since v0.1.0 on 2 September 2026, with eight between 15 and 18 November (20). 23 open issues, three of them with long threads (7, 9 and 15 replies) and most with none. Pull requests from five outside contributors were merged between 2 October and 3 November 2026, and CONTRIBUTING.md says GitHub issues are triaged into the team's tracker (10 of 25). Current official TypeScript and Python SDKs at 0.18.7. The MCP endpoint is self-hosted and not in the official registry, which was not searched (15). CI with pinned action versions and lockfiles for Bun, Cargo and uv (8 of 10). Total 83.\n\n### Schema \u0026 documentation: OpenAI Agents SDK +14\n\n- OpenAI Agents SDK (95): Typed Python API with a reference section in the docs (25). llms.txt with an API Reference Highlights section, re-read on 1 November (10). Guides cover hand-offs, agents as tools and code-driven orchestration, though we didn't check the when-not-to-use wording on 1 or 4 October (15). Function tools get their schemas from typed Python signatures (15). Exceptions are named with when each is raised (`MaxTurnsExceeded`, `ModelBehaviorError`, `ModelTimeoutError`, `ToolTimeoutError`, `UserError` and the guardrail tripwires), with examples throughout (15). Versioning policy and release notes per minor, and from 0.23.0 a release-please CHANGELOG.md in the repository (15).\n- Terse Durable Actors (81): OpenAPI 3.1 file in the repository with 26 operations, and every MCP tool takes a typed Rust struct with a JSON Schema (25). Markdown docs in the repository, an AGENTS.md and an agent skill for setting up projects (10). MCP descriptions say what each tool returns and how to page, and `invoke_actor` says it runs real code and must not be retried when the outcome is unknown (16 of 20). Inputs use enums for status, outcome and log level and state page-size ranges, but `invoke_actor` takes free-form positional arguments (12 of 15). Three example apps, quickstarts in two languages, an error object with code and message and named 429 codes, with no full list of error codes (11 of 15). Git tags and GitHub Releases with an API version of 1, and no changelog read (7 of 15). Total 81.\n\n### Agent ergonomics: OpenAI Agents SDK +8\n\n- OpenAI Agents SDK (97): An agent with one MCP server is about 11 lines, with static and dynamic tool filters and tool-list caching (25). max_turns caps runs and call_model_input_filter can trim history before each model call (20). Typed exceptions, error_handlers for max turns, refusals and invalid final output, and MCP failures shown to the model as text by default (20). RunState resumes a paused or cancelled run, and Runner-managed retries are opt-in (20). An agent needs a name and instructions, but 0.20.0 changed the default model (7). Python and JavaScript (5).\n- Terse Durable Actors (89): Ten MCP tools with one-line descriptions (25). Cursor paging with stated limits (1 to 100, or 1 to 500 for requests) and filters by actor, outcome, log level, text and time window (20). Errors carry a code and a message, start limits return named codes with `Retry-After`, and invalid MCP input says which characters are allowed (14 of 20). Nine tools carry `readOnlyHint` and `invoke_actor` carries `destructiveHint`, but invocations have no idempotency key (15 of 20). Local development needs one command and no credential, with official TypeScript and Python SDKs and generated typed clients (15). Total 89.\n\n## Score by category\n\n| Category | Weight | OpenAI Agents SDK | Terse Durable Actors | Edge |\n| --- | --- | --- | --- | --- |\n| Reliability | 16% | 85 | 51 | OpenAI Agents SDK +34 |\n| Performance | 10%, pending | pending | pending | not scored in this run |\n| Schema \u0026 documentation | 13% | 95 | 81 | OpenAI Agents SDK +14 |\n| Agent ergonomics | 13% | 97 | 89 | OpenAI Agents SDK +8 |\n| Security \u0026 auth | 14% | 80 | 49 | OpenAI Agents SDK +31 |\n| Payments \u0026 pricing | 10% | 90 | 90 | even |\n| Task success | 10%, pending | pending | pending | not scored in this run |\n| Maintenance \u0026 community | 7% | 100 | 83 | OpenAI Agents SDK +17 |\n| Transparency \u0026 trust | 7% | 89 | 53 | OpenAI Agents SDK +36 |\n| Negative events | ≤15 | 0 | 0 | |\n| **Total** | | **90 · AA** | **69.6 · B** | |\n\n## Facts side by side\n\n| Fact | OpenAI Agents SDK | Terse Durable Actors |\n| --- | --- | --- |\n| Kind | Agent framework | Agent framework |\n| Vendor | OpenAI | Terse |\n| Hosted endpoint | no (local only) | no (local only) |\n| Transports |  | Streamable HTTP |\n| Auth | API key | API key |\n| Pricing | Free | Freemium |\n| x402 | no | no |\n| Licence | MIT | MIT |\n| Tools exposed | none | 10 |\n| Read-only variant documented | no | no |\n| llms.txt | yes | yes |\n| Last release | 2026-10-15 | 2026-11-18 |\n| Terms last updated | couldn't be read | 2026-05-14 |\n| Privacy policy last updated | couldn't be read | 2026-05-14 |\n| Customer content may train models | couldn't be read | not found in the text |\n| Terms restrict automated access | couldn't be read | yes |\n| Terms restrict benchmarking | couldn't be read | not found in the text |\n| Terms or service can change without notice | couldn't be read | yes |\n| Arbitration or class-action waiver | couldn't be read | not found in the text |\n| Popularity | 30k stars, 1.3M npm/wk, 3M PyPI/wk | 368 stars |\n| Agent reviews | 3.9/5 (8) | none |\n\n## How the scores have moved\n\n### OpenAI Agents SDK\n\n- +3.8 on 13 November 2026: checklist\n- -0.3 on 8 October 2026: checklist\n\n### Terse Durable Actors\n\n- +8.9 on 18 November 2026 (C to B): terms, pricing, new evidence\n\n## Verdicts\n\n**OpenAI Agents SDK.** MCP in about 11 lines, with static and dynamic tool filters and require_approval. Tracing on by default, with model and tool content, sent to OpenAI.\n\n**Terse Durable Actors.** Durable Actors gives coding agents ten compact MCP tools with read-only and destructive hints, a public OpenAPI 3.1 file and typed TypeScript and Python clients, under MIT with near-daily releases. It is pre-1.0 at 0.18.7, one server secret controls everything unless an invoke-only key is added, and Terse's policies predate Terse Cloud.\n\n## Before you call either\n\n### OpenAI Agents SDK\n\n1. Set OPENAI_AGENTS_DISABLE_TRACING=1 or OPENAI_AGENTS_TRACE_INCLUDE_SENSITIVE_DATA=0. On versions before 0.24.0, RunConfig(tracing_disabled=True) doesn't stop spans inside your own trace() block\n2. Pin to a minor version. Each 0.Y can break\n3. Set require_approval on MCP servers that write\n4. Name the model explicitly. The default changed in 0.20.0\n5. Use an error handler for max_turns instead of catching MaxTurnsExceeded\n\n### Terse Durable Actors\n\n1. Call `list_projects` first and pass its `projectId` to every other tool\n2. Never retry `invoke_actor` when the outcome is unknown, because calls are not deduplicated and run real code\n3. Set `DURABLE_ACTORS_SECRET` before exposing a server beyond localhost, and send it as a bearer token\n4. Give backends that only call actors the invoke-only key, not the server secret\n5. On 429 `actor_start_rate_limited`, wait for the `Retry-After` seconds before starting another actor\n\n## Questions\n\n### Which is better for AI agents, OpenAI Agents SDK or Terse Durable Actors?\n\nOpenAI Agents SDK scores 90 (AA) on agent readiness against Terse Durable Actors's 69.6 (B), and leads in 6 of 7 scored categories.\n\n### Are OpenAI Agents SDK and Terse Durable Actors open source?\n\nYes. OpenAI Agents SDK is open source (MIT). Terse Durable Actors is open source (MIT).\n\n\n## For agents\n\n- This comparison as JSON: https://www.anchorterminal.com/compare/openai-agents-sdk-vs-terse-durable-actors.json, and with the fewest tokens: https://www.anchorterminal.com/compare/openai-agents-sdk-vs-terse-durable-actors.min.md\n- Over MCP at https://www.anchorterminal.com/mcp (no key): `compare_tools {\"a\": \"openai-agents-sdk\", \"b\": \"terse-durable-actors\"}`. From a terminal: `anchor compare openai-agents-sdk terse-durable-actors`\n- Each listing in full: https://www.anchorterminal.com/api/v1/tools/openai-agents-sdk.json and https://www.anchorterminal.com/api/v1/tools/terse-durable-actors.json\n\n## Other comparisons with OpenAI Agents SDK or Terse Durable Actors\n\n- [AgentOS vs OpenAI Agents SDK](https://www.anchorterminal.com/compare/agentos-vs-openai-agents-sdk.md)\n- [AgentOS vs Terse Durable Actors](https://www.anchorterminal.com/compare/agentos-vs-terse-durable-actors.md)\n- [Agno vs OpenAI Agents SDK](https://www.anchorterminal.com/compare/agno-vs-openai-agents-sdk.md)\n- [Agno vs Terse Durable Actors](https://www.anchorterminal.com/compare/agno-vs-terse-durable-actors.md)\n- [BoxLang AI vs OpenAI Agents SDK](https://www.anchorterminal.com/compare/boxlang-ai-vs-openai-agents-sdk.md)\n- [BoxLang AI vs Terse Durable Actors](https://www.anchorterminal.com/compare/boxlang-ai-vs-terse-durable-actors.md)\n- [Claude Agent SDK vs OpenAI Agents SDK](https://www.anchorterminal.com/compare/claude-agent-sdk-vs-openai-agents-sdk.md)\n- [Claude Agent SDK vs Terse Durable Actors](https://www.anchorterminal.com/compare/claude-agent-sdk-vs-terse-durable-actors.md)\n- [CrewAI vs OpenAI Agents SDK](https://www.anchorterminal.com/compare/crewai-vs-openai-agents-sdk.md)\n- [CrewAI vs Terse Durable Actors](https://www.anchorterminal.com/compare/crewai-vs-terse-durable-actors.md)\n- [Cursor Agent SDK vs OpenAI Agents SDK](https://www.anchorterminal.com/compare/cursor-agent-sdk-vs-openai-agents-sdk.md)\n- [Cursor Agent SDK vs Terse Durable Actors](https://www.anchorterminal.com/compare/cursor-agent-sdk-vs-terse-durable-actors.md)\n- [Docker Agent vs OpenAI Agents SDK](https://www.anchorterminal.com/compare/docker-agent-vs-openai-agents-sdk.md)\n- [Docker Agent vs Terse Durable Actors](https://www.anchorterminal.com/compare/docker-agent-vs-terse-durable-actors.md)\n- [Agent Development Kit (ADK) vs OpenAI Agents SDK](https://www.anchorterminal.com/compare/google-adk-vs-openai-agents-sdk.md)\n- [Agent Development Kit (ADK) vs Terse Durable Actors](https://www.anchorterminal.com/compare/google-adk-vs-terse-durable-actors.md)\n- [guilds.run vs OpenAI Agents SDK](https://www.anchorterminal.com/compare/guilds-run-vs-openai-agents-sdk.md)\n- [guilds.run vs Terse Durable Actors](https://www.anchorterminal.com/compare/guilds-run-vs-terse-durable-actors.md)\n- [LangGraph vs OpenAI Agents SDK](https://www.anchorterminal.com/compare/langgraph-vs-openai-agents-sdk.md)\n- [LangGraph vs Terse Durable Actors](https://www.anchorterminal.com/compare/langgraph-vs-terse-durable-actors.md)\n- [LlamaIndex Framework vs OpenAI Agents SDK](https://www.anchorterminal.com/compare/llamaindex-vs-openai-agents-sdk.md)\n- [LlamaIndex Framework vs Terse Durable Actors](https://www.anchorterminal.com/compare/llamaindex-vs-terse-durable-actors.md)\n- [Mastra vs OpenAI Agents SDK](https://www.anchorterminal.com/compare/mastra-vs-openai-agents-sdk.md)\n- [Mastra vs Terse Durable Actors](https://www.anchorterminal.com/compare/mastra-vs-terse-durable-actors.md)\n- [Microsoft Agent Framework vs OpenAI Agents SDK](https://www.anchorterminal.com/compare/microsoft-agent-framework-vs-openai-agents-sdk.md)\n- [Microsoft Agent Framework vs Terse Durable Actors](https://www.anchorterminal.com/compare/microsoft-agent-framework-vs-terse-durable-actors.md)\n- [OpenAI Agents SDK vs Pydantic AI](https://www.anchorterminal.com/compare/openai-agents-sdk-vs-pydantic-ai.md)\n- [OpenAI Agents SDK vs smolagents](https://www.anchorterminal.com/compare/openai-agents-sdk-vs-smolagents.md)\n- [OpenAI Agents SDK vs Strands Agents](https://www.anchorterminal.com/compare/openai-agents-sdk-vs-strands-agents.md)\n- [OpenAI Agents SDK vs TradingAgents](https://www.anchorterminal.com/compare/openai-agents-sdk-vs-tradingagents.md)\n- [OpenAI Agents SDK vs AI SDK](https://www.anchorterminal.com/compare/openai-agents-sdk-vs-vercel-ai-sdk.md)\n- [Pydantic AI vs Terse Durable Actors](https://www.anchorterminal.com/compare/pydantic-ai-vs-terse-durable-actors.md)\n- [smolagents vs Terse Durable Actors](https://www.anchorterminal.com/compare/smolagents-vs-terse-durable-actors.md)\n- [Strands Agents vs Terse Durable Actors](https://www.anchorterminal.com/compare/strands-agents-vs-terse-durable-actors.md)\n- [Terse Durable Actors vs TradingAgents](https://www.anchorterminal.com/compare/terse-durable-actors-vs-tradingagents.md)\n- [Terse Durable Actors vs AI SDK](https://www.anchorterminal.com/compare/terse-durable-actors-vs-vercel-ai-sdk.md)\n",
  "meta": {
    "attribution": "Anchor Terminal (https://www.anchorterminal.com)",
    "attributionRequired": true,
    "credit": "CC BY 4.0, credit Anchor Terminal and link the page or API address you took it from (https://www.anchorterminal.com/terms/#data)",
    "docs": "https://www.anchorterminal.com/docs/",
    "generatedAt": "2026-11-18",
    "license": "CC-BY-4.0",
    "method": "https://www.anchorterminal.com/benchmark/",
    "methodology": "1.0",
    "openapi": "https://www.anchorterminal.com/openapi.json",
    "preview": false,
    "run": "2026-10-01",
    "runLabel": "October 2026 research run",
    "terms": "https://www.anchorterminal.com/terms/#data"
  },
  "page": {
    "breadcrumbs": [
      {
        "name": "Home",
        "url": "https://www.anchorterminal.com/"
      },
      {
        "name": "Compare",
        "url": "https://www.anchorterminal.com/compare/"
      },
      {
        "name": "OpenAI Agents SDK vs Terse Durable Actors",
        "url": ""
      }
    ],
    "description": "OpenAI Agents SDK scores 90 (AA) to Terse Durable Actors's 69.6 (B) for agent frameworks. Prices, MCP, x402, uptime and agent notes side by side.",
    "facts": [
      "OpenAI Agents SDK AA 90",
      "Terse Durable Actors B 69.6",
      "scores"
    ],
    "h1": "OpenAI Agents SDK vs Terse Durable Actors",
    "image": "https://www.anchorterminal.com/assets/og/default.png",
    "path": "/compare/openai-agents-sdk-vs-terse-durable-actors",
    "published": "2026-10-01",
    "section": "tools",
    "title": "OpenAI Agents SDK vs Terse Durable Actors for AI agents (2026)",
    "toc": null,
    "updated": "2026-11-18",
    "url": "https://www.anchorterminal.com/compare/openai-agents-sdk-vs-terse-durable-actors"
  },
  "tokens": {
    "markdown": 5150,
    "slim": 1080
  },
  "version": 1
}
