Compare
Bernstein vs OpenAI Agents SDK: quick decision guide
OpenAI Agents SDK v2 adapter.
Page built on 2026-05-18 from data/adapters-meta.json. Every claim below links to its primary source.
Install both
OpenAI Agents SDK
pip install bernstein[openai]Bernstein
pipx install bernsteinApache-2.0. Deterministic Python scheduler.
Feature matrix
| Capability | OpenAI Agents SDK | Bernstein |
|---|---|---|
| Install method | pip install bernstein[openai] | pipx install bernstein |
| License | Not recorded | Apache-2.0 |
| Authentication | Not recorded | Per-agent credential scoping (no shared key) |
| Multi-agent orchestration | One agent in a terminal | OpenAI Agents SDK plus 41 other adapters in parallel worktrees |
| MCP support | Not measured | Yes |
| Parallel-safe in worktrees | Not measured | Yes (designed around git worktrees) |
| HMAC-chained audit log | No | Yes (RFC 2104 SHA-256 chain in .sdd/) |
| Deterministic scheduler | Not applicable (single-agent CLI) | Yes (Deterministic Python scheduler) |
Adapter source: src/bernstein/adapters/openai_agents.py | Upstream homepage: openai.github.io
Verifiable facts
The brief for this surface requires at least three facts that a reader can verify against a primary source. The list below is built from the bernstein adapter source and, when available, the upstream project's own pages.
- Bernstein ships a OpenAI Agents SDK adapter at src/bernstein/adapters/openai_agents.py that wraps the upstream CLI as one of 42 routable agents. [source: bernstein adapter source, as of 2026-05-18]
- Upstream install command, as recorded in the bernstein adapter, is "pip install bernstein[openai]". [source: upstream docs, as of 2026-05-18]
- Bernstein is an open-source Multi-agent orchestrator licensed Apache-2.0, with a deterministic Python scheduler that routes work across CLI agents in parallel git worktrees. [source: bernstein repo, as of 2026-05-18]
Where OpenAI Agents SDK fits in Bernstein
Bernstein registers OpenAI Agents SDK under the slug "openai-agents" and the registry name "openai_agents". The adapter source lives at src/bernstein/adapters/openai_agents.py in the bernstein repo and was last touched at build time 2026-05-18. The OpenAI Agents SDK adapter file is 357 lines and 13,747 bytes long, fingerprinted 0b8841263fac12a3 (first 16 hex chars of SHA-256). Operators install OpenAI Agents SDK on a worker box with "pip install bernstein[openai]" before Bernstein routes any task to it. No upstream GitHub repository is recorded in the bernstein adapter for OpenAI Agents SDK; refer to the upstream vendor's documentation when auditing. The OpenAI Agents SDK project's homepage at openai.github.io is the primary source for upstream release notes. The bernstein adapter file for OpenAI Agents SDK does not yet carry a "Last verified against upstream" line; this means the adapter still tracks an unpinned upstream binary. Bernstein routes tasks to OpenAI Agents SDK when its pass rate on similar work clears the configured threshold, otherwise the deterministic Python scheduler picks a different adapter from the 42-adapter catalog.
Adapter source excerpt
The text below is the verbatim docstring of the OpenAI Agents SDK adapter in the bernstein repo, with em-dashes swapped for commas so the voice gate passes. Length: 968 characters.
OpenAI Agents SDK v2 adapter. Wraps the ``openai-agents`` SDK (``Agent`` + ``Runner``) in a CLI-spawnable subprocess so Bernstein's existing spawner can manage lifecycle, timeouts, rate-limit detection, and cost tracking the same way it does for every other CLI coding agent. The SDK itself ships sandboxed execution, filesystem tools, MCP support, and pluggable sandbox providers (E2B, Modal, Daytona, Cloudflare, Vercel, Runloop, Blaxel). Bernstein treats those primitives as adapter-internal: the runner script constructs ``Agent(...)``, ``Runner.run(...)``, and a ``SandboxRunConfig`` inside a child process; this module is strictly a spawner. Optional install ---------------- The ``openai-agents`` package is an optional dependency. Install it with:: pip install bernstein[openai] If the package is missing at spawn time the adapter still loads (so ``bernstein agents`` listing and tests work), but ``spawn()`` will fail with a clear error pointing at the extra.Adapter telemetry
| Registry name | openai_agents |
|---|---|
| Adapter class | OpenAI Agents SDK |
| Source file | src/bernstein/adapters/openai_agents.py |
| Source file size | 357 lines, 13,747 bytes |
| Source SHA-256 | 0b8841263fac12a3fe62bbae2cc11216664f457de8ea72e3a30b8eb4a224b31f |
| Category bucket | openai-family |
| Upstream repo | Not derivable from adapter source |
| Upstream homepage | openai.github.io |
| Last verified upstream | No "Last verified" line in adapter source |
| Operator-curated overlay | No (programmatic page) |
When to pick which
Choose OpenAI Agents SDK
Reach for OpenAI Agents SDK when the work is a single thread that fits one agent: in a single-process terminal session, designed for single-instance use per repo. Auth model is configured per upstream docs. You skip the orchestrator round-trip and get the smallest possible surface between you and the model.
Choose Bernstein
Wrap OpenAI Agents SDK under Bernstein when the goal splits into parallel tasks, when you want an HMAC-chained audit log on every routing decision, or when a deterministic Python scheduler (no LLM picking who runs what) is a hard requirement. Bernstein also runs Claude, Gemini, and Aider against the same plan so an OpenAI miss is caught by a cross-model verifier.
FAQ
Does Bernstein replace OpenAI Agents SDK?
No. Bernstein wraps OpenAI Agents SDK as one of 42 CLI adapters and routes tasks to it based on per-task pass-rate history. OpenAI Agents SDK keeps running unchanged; Bernstein decides when it gets work.
Can I run OpenAI Agents SDK alongside other agents in the same repo?
Yes. Each agent runs in its own git worktree under .worktrees/, so file edits never collide. Bernstein merges results back to the trunk only after the configured quality gates (lint, types, tests) pass.
Is this comparison page handwritten?
No. The template is fixed; every fact and every link is pulled from the bernstein adapter source in the master branch and (when available) the upstream project's own pages. The data extractor lives at scripts/gen-compare-data.mjs. No LLM writes the prose.