{"collection":{"slug":"frontier-ai-labs","name":"Frontier AI Labs"},"digest":{"id":"cwd_4vSVHv85S8Rz4oAvzNNOA","weekStart":"2026-10-26","title":"GPT-6 goes to the life sciences as the SDKs converge on agents","intro":"OpenAI put GPT-6 behind biology-specific safeguards with GPT-6 Rosalind Research, rolling it out across Work, Codex, and the API for approved life sciences teams, while Codex and the OpenAI SDKs spent the week building out agent orchestration, and Claude Code hardened its permission model.","body":"### GPT-6 goes to the bench\n\nThe week's biggest story is OpenAI's decision that its frontier reasoning model needed a domain-specific variant. [GPT-6 Rosalind Research](/release/rel_38Gl2Tn-hz5ZWf5wvjjAG-openai-launches-gpt-6-rosalind-research-for-multistep-life-sciences-work) brings GPT-6 reasoning to multistep life sciences research with safeguards built for biological work — and access is gated accordingly: organizations and individual users need a separate Research approval, and prior Discovery approval doesn't count. Enterprise customers get it through a [two-week Admin Preview that starts off by default](/release/rel_0U_gh2KVd7InoN_sa_E9Y-chatgpt-enterprise-edu-rolls-out-gpt-6-rosalind-research-in-two-week-admin), while approved Business customers can pick it up directly in [Work and Codex](/release/rel_klAXbgyARrcTjU4U3Y2mN-chatgpt-business-rolls-out-gpt-6-rosalind-research-to-approved-life-sciences). GPT-5.5 Rosalind Research stays available for existing workflows throughout, so nothing breaks for teams that don't migrate yet.\n\nAnthropic is telling a quieter version of the same story: its scientists reported [early progress from deep research partnerships with six drug developers](/release/rel_9N2KJq9et8LGXygmZc9AR-anthropic-reports-early-progress-on-accelerating-drug-development-with-claude), framing Claude as a tool for accelerating medicine discovery rather than a product launch.\n\n### Agents become the unit of work in the SDKs\n\nThe OpenAI SDKs converged on the same set of primitives across languages this week. The Node SDK led with [scheduled agent automations and delegation controls](/release/rel_wQr5aYJ1ee0uoaaGBVDLs-openai-node-sdk-v7-44-0-adds-agent-automations-and-delegation-controls) — a `client.beta.agents.automations` namespace plus an automations agent-tool variant — and a breaking change to Live Decision events that will catch anyone reading `item.content` or `item.role` directly. It followed with [agent session review and tool settings, context compaction, and experimental time awareness](/release/rel_CS8H8QNz1c4vqpSwgC82Z-openai-node-sdk-v7-46-0-adds-agent-session-review-and-tool-settings), and added [typed Responses WebSocket warmup](/release/rel_HVVeaMYyBgZScWsPV_o7H-openai-node-sdk-v7-47-0-adds-typed-responses-websocket-warmup). The Python SDK landed the same foundation — [delegation controls and prompt cache lifetimes](/release/rel_VnaN3lCvKUoYe6HvI-aKF-openai-python-sdk-v3-38-0-adds-delegation-controls-and-prompt-cache-lifetimes) first, then [response tool usage models and Realtime typed references](/release/rel_8_kg7w__zWWj4yPt_Znbu-openai-python-sdk-v3-41-0-adds-response-tool-usage-models), followed by its own [agent context compaction and session review settings](/release/rel_T52RUsDfAhuNiI4L6brYc-openai-python-sdk-v3-42-0-adds-agent-context-compaction-and-session-review). Go tracked in parallel, adding [delegation controls and cache lifetimes](/release/rel_l0xTEcPaRgui6oYen6wn8-openai-go-sdk-v3-87-0-adds-delegation-controls-and-prompt-cache-lifetimes) and [agent compaction, session review, and SIP DTMF delivery events](/release/rel_qPJlshH54xYJRys1-X9Nf-openai-go-sdk-v3-91-0-adds-agent-context-compaction-and-session-review-settings), with [image and web search usage reporting](/release/rel_Z6b__RNr3Jum2zYYFG9yO-openai-go-sdk-v3-90-0-adds-usage-tracking-for-image-and-web-search-in-responses) in between. Prompt-cache TTL is now something callers set explicitly rather than inherit: [Python exposed the options](/release/rel_y56Y3OF-1EIPx47aYO2zZ-openai-python-sdk-v3-40-0-adds-prompt-cache-breakpoint-ttl-options), [Go too](/release/rel_ql4SVN9N0e2-k-oT5ESUb-openai-go-sdk-v3-89-0-adds-prompt-cache-breakpoint-ttl-options), and [Node clarified the breakpoint semantics](/release/rel_RJ9iARUNFADPfI6K0_mAh-openai-node-sdk-v7-45-1-clarifies-prompt-cache-breakpoint-ttl-and-adds-google).\n\nAll three SDKs also picked up the [gpt-6.1-terra model identifier](/release/rel_Cz1ewjydlKZQlfqzTLNcv-openai-node-sdk-v7-45-0-adds-gpt-6-1-terra-model-identifier) — [the same addition in Python](/release/rel_bLz-dZtlFKjskB2U-NTAU-openai-python-sdk-v3-39-0-adds-gpt-6-1-terra-model-identifier) and [in Go](/release/rel_nENFj8Nrv_IbFIK82tUmU-openai-go-sdk-v3-88-0-adds-gpt-6-1-terra-model-identifier) — and Go shipped [a Live WebSocket guide](/release/rel_WmstzSLoL9UObU26VyKYj-openai-go-sdk-v3-91-1-adds-live-websocket-guide). Python wrapped up with [WebSocket behavior preservation and realtime auth validation](/release/rel_D0U8tKGAXFWX6y5qOP4nu-openai-python-sdk-v3-42-1-preserves-responses-websocket-behavior-and-validates).\n\n### Codex and Claude Code harden the local agent loop\n\nCodex CLI added a [built-in Vertex AI provider and GPT-6.1 Terra in the model catalogs](/release/rel_ylZHV7lEyapGlaXXkrk29-codex-cli-0-165-0-adds-vertex-ai-provider-and-gpt-6-1-terra), turned on memory v2 and mid-conversation reasoning-effort changes by default, and closed a marketplace URL validation gap that could send Git to a different host than the one checked. It then extended Terra to the [bundled, Vertex, and Bedrock catalogs](/release/rel_8BkJacLBZaovUaPmsX-qp-codex-cli-0-166-0-adds-gpt-6-1-terra-and-keyboard-driven-transcript-selection) and gave transcripts keyboard-driven selection with integrated search.\n\nClaude Code spent the week on trust boundaries and reliability. A [write-block setting that refuses edits outside working directories in every permission mode](/release/rel_GVnwz7Q4leUVNJVfqLNdq-claude-code-v2-1-314-adds-write-block-setting-and-fixes-turn-after-turn-effort) came alongside idle compaction for long VS Code sessions, and a fix so the Write tool [can no longer clobber a large CLAUDE.md after seeing only its opening lines](/release/rel_mapQ4E72ZAef5Bh48WVjA-claude-code-v2-1-312-stops-write-from-clobbering-large-claude-md-files). Startup freezes from pathological CLAUDE.md list items and [context-limit stalls that now compact instead of stopping](/release/rel_SUgZu2mcsNgA4M9QNDQ0a-claude-code-v2-1-311-fixes-startup-freezes-and-1m-context-billing-stops) rounded out performance work. Opus 4.8 fast mode is [deprecated with removal set for November 30](/release/rel_RtkLb4jV3d50qor9FjLYd-claude-code-v2-1-316-deprecates-opus-4-8-fast-mode-adds-subagent-effort-control) — migrate to Opus 5.5 to keep the speedup — and Claude Code gained subagent effort control and pooled group budgets for the apps gateway.\n\n### Product surface and platform plumbing\n\nChatGPT itself got more conversational infrastructure than model news: [texting your dot via iMessage or RCS](/release/rel_q7lKmiDVrUoX6VmnOilx_-chatgpt-rolls-out-beta-texting-for-your-dot-to-pro-users-in-the-us) began a U.S. beta for Pro users, [account switching arrived on iOS and Android](/release/rel_UwevlxtYDP84n9Htu4AmO-chatgpt-rolls-out-account-switching-on-ios-and-android), [iCloud Mail and Calendar became connectable apps](/release/rel_DlR_k1dbz6AnMO2uX90Bk-chatgpt-connects-icloud-mail-and-calendar-for-email-and-schedule-lookups), and Free and Go users got [more concise, less flattering GPT-6 replies](/release/rel_BIKlM7XFvprZgI3HgREmF-openai-makes-gpt-6-in-chatgpt-more-concise-and-less-flattering-for-free-and-go). On the enterprise side, automatic Instant-to-thinking switching was [retired in favor of explicit model-picker choices](/release/rel_kNS1ao-uEz3qakFgQJch--chatgpt-enterprise-edu-retires-automatic-instant-to-thinking-model-switching). Usage got more generous and more measurable: Pro and Business Premium plans [doubled their Work and Codex allowances](/release/rel_xlxjl8JvcdJMwuZqP4uYq-chatgpt-doubles-work-and-codex-usage-on-pro-and-business-premium-plans), eligible subscribers can claim [up to $500 a month in promotional API credits](/release/rel_yzaoey9-A32Gja1tXigmi-openai-api-offers-chatgpt-pro-and-business-subscribers-up-to-500-monthly-in-free) — flagged for [Business workspaces](/release/rel_JEHWGD44eDGlqN-vcRnVL-chatgpt-business-adds-promotional-api-credits-for-eligible-workspaces) and [Pro and Business subscribers](/release/rel_S9kt-lYrjrwjRvgDw42RV-chatgpt-pro-and-business-subscribers-can-activate-promotional-api-credits) separately — and the Claude Platform opened a [Usage and Cost Analytics API in beta](/release/rel_w5k4pmjpV1ne9RsUqdZe2-claude-platform-launches-usage-and-cost-analytics-api-in-beta) while [raising the Messages API request ceiling to 128 MB](/release/rel_cUdw9pwBkz_45EydzHTNr-claude-platform-raises-messages-api-max-request-size-to-128-mb). A [Model Spec reorganized around five core values](/release/rel_dRmjmW0FqWHjcIfzn0yRj-openai-reorganizes-model-spec-around-five-core-values) rounds out the week's policy-facing changes.","releaseIds":["rel_38Gl2Tn-hz5ZWf5wvjjAG","rel_0U_gh2KVd7InoN_sa_E9Y","rel_klAXbgyARrcTjU4U3Y2mN","rel_9N2KJq9et8LGXygmZc9AR","rel_wQr5aYJ1ee0uoaaGBVDLs","rel_CS8H8QNz1c4vqpSwgC82Z","rel_HVVeaMYyBgZScWsPV_o7H","rel_VnaN3lCvKUoYe6HvI-aKF","rel_8_kg7w__zWWj4yPt_Znbu","rel_T52RUsDfAhuNiI4L6brYc","rel_l0xTEcPaRgui6oYen6wn8","rel_qPJlshH54xYJRys1-X9Nf","rel_Z6b__RNr3Jum2zYYFG9yO","rel_y56Y3OF-1EIPx47aYO2zZ","rel_ql4SVN9N0e2-k-oT5ESUb","rel_RJ9iARUNFADPfI6K0_mAh","rel_Cz1ewjydlKZQlfqzTLNcv","rel_bLz-dZtlFKjskB2U-NTAU","rel_nENFj8Nrv_IbFIK82tUmU","rel_WmstzSLoL9UObU26VyKYj","rel_D0U8tKGAXFWX6y5qOP4nu","rel_ylZHV7lEyapGlaXXkrk29","rel_8BkJacLBZaovUaPmsX-qp","rel_GVnwz7Q4leUVNJVfqLNdq","rel_mapQ4E72ZAef5Bh48WVjA","rel_SUgZu2mcsNgA4M9QNDQ0a","rel_RtkLb4jV3d50qor9FjLYd","rel_q7lKmiDVrUoX6VmnOilx_","rel_UwevlxtYDP84n9Htu4AmO","rel_DlR_k1dbz6AnMO2uX90Bk","rel_BIKlM7XFvprZgI3HgREmF","rel_kNS1ao-uEz3qakFgQJch-","rel_xlxjl8JvcdJMwuZqP4uYq","rel_yzaoey9-A32Gja1tXigmi","rel_JEHWGD44eDGlqN-vcRnVL","rel_S9kt-lYrjrwjRvgDw42RV","rel_w5k4pmjpV1ne9RsUqdZe2","rel_cUdw9pwBkz_45EydzHTNr","rel_dRmjmW0FqWHjcIfzn0yRj"],"releaseCount":45,"generatedAt":"2026-11-02T06:17:04.127Z","releases":[{"id":"rel_38Gl2Tn-hz5ZWf5wvjjAG","title":"OpenAI launches GPT-6 Rosalind Research for multistep life sciences work","path":"/release/rel_38Gl2Tn-hz5ZWf5wvjjAG-gpt-6-rosalind-research-opens-to-approved-customers","url":"https://help.openai.com/en/articles/9624314-model-release-notes#gpt-6-rosalind-research","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"models","name":"Models"},"importance":5,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-30T00:00:00.000Z"},{"id":"rel_0U_gh2KVd7InoN_sa_E9Y","title":"ChatGPT Enterprise & Edu rolls out GPT-6 Rosalind Research in two-week Admin Preview","path":"/release/rel_0U_gh2KVd7InoN_sa_E9Y-gpt-6-rosalind-research-ships-in-off-by-default-admin-preview","url":"https://help.openai.com/en/articles/10128477-chatgpt-enterprise-edu-release-notes#gpt-6-rosalind-research","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt-enterprise-edu","name":"ChatGPT Enterprise & Edu"},"importance":4,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-30T00:00:00.000Z"},{"id":"rel_klAXbgyARrcTjU4U3Y2mN","title":"ChatGPT Business rolls out GPT-6 Rosalind Research to approved life sciences customers","path":"/release/rel_klAXbgyARrcTjU4U3Y2mN-gpt-6-rosalind-research-limited-rollout-for-life-sciences","url":"https://help.openai.com/en/articles/11391654-chatgpt-business-release-notes#gpt-6-rosalind-research","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt-business","name":"ChatGPT Business"},"importance":4,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-30T00:00:00.000Z"},{"id":"rel_9N2KJq9et8LGXygmZc9AR","title":"Anthropic reports early progress on accelerating drug development with Claude","path":"/release/rel_9N2KJq9et8LGXygmZc9AR-early-drug-development-progress-with-claude","url":"https://www.anthropic.com/research/early-progress-drug-development","org":{"slug":"anthropic","name":"Anthropic","avatarUrl":"https://media.releases.sh/orgs/anthropic.png","githubHandle":"anthropics"},"product":null,"importance":3,"composition":null,"publishedAt":"2026-10-27T00:00:00.000Z"},{"id":"rel_wQr5aYJ1ee0uoaaGBVDLs","title":"OpenAI Node SDK v7.44.0 adds agent automations and delegation controls","path":"/release/rel_wQr5aYJ1ee0uoaaGBVDLs-agent-automations-api-added-live-decision-item-types-break","url":"https://github.com/openai/openai-node/releases/tag/v7.44.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":4,"composition":{"bugs":0,"features":3,"enhancements":1},"publishedAt":"2026-10-26T15:31:46.000Z"},{"id":"rel_CS8H8QNz1c4vqpSwgC82Z","title":"OpenAI Node SDK v7.46.0 adds agent session review and tool settings","path":"/release/rel_CS8H8QNz1c4vqpSwgC82Z-agent-session-review-and-tool-settings-external-key-management-added","url":"https://github.com/openai/openai-node/releases/tag/v7.46.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":3,"composition":{"bugs":1,"features":9,"enhancements":0},"publishedAt":"2026-10-31T00:27:06.000Z"},{"id":"rel_HVVeaMYyBgZScWsPV_o7H","title":"OpenAI Node SDK v7.47.0 adds typed Responses WebSocket warmup","path":"/release/rel_HVVeaMYyBgZScWsPV_o7H-responses-websocket-warmup-now-typed","url":"https://github.com/openai/openai-node/releases/tag/v7.47.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":2,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-31T04:46:03.000Z"},{"id":"rel_VnaN3lCvKUoYe6HvI-aKF","title":"OpenAI Python SDK v3.38.0 adds delegation controls and prompt cache lifetimes","path":"/release/rel_VnaN3lCvKUoYe6HvI-aKF-delegation-controls-and-prompt-cache-lifetimes-added","url":"https://github.com/openai/openai-python/releases/tag/v3.38.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":3,"composition":{"bugs":5,"features":2,"enhancements":0},"publishedAt":"2026-10-26T16:35:03.000Z"},{"id":"rel_8_kg7w__zWWj4yPt_Znbu","title":"OpenAI Python SDK v3.41.0 adds response tool usage models","path":"/release/rel_8_kg7w__zWWj4yPt_Znbu-response-tool-usage-models-agent-sessions-filter-by-metadata","url":"https://github.com/openai/openai-python/releases/tag/v3.41.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":3,"composition":{"bugs":3,"features":4,"enhancements":0},"publishedAt":"2026-10-28T22:26:54.000Z"},{"id":"rel_T52RUsDfAhuNiI4L6brYc","title":"OpenAI Python SDK v3.42.0 adds agent context compaction and session review settings","path":"/release/rel_T52RUsDfAhuNiI4L6brYc-agent-context-compaction-and-session-review-settings-added","url":"https://github.com/openai/openai-python/releases/tag/v3.42.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":3,"composition":{"bugs":1,"features":9,"enhancements":0},"publishedAt":"2026-10-31T00:26:41.000Z"},{"id":"rel_l0xTEcPaRgui6oYen6wn8","title":"OpenAI Go SDK v3.87.0 adds delegation controls and prompt cache lifetimes","path":"/release/rel_l0xTEcPaRgui6oYen6wn8-delegation-controls-prompt-cache-lifetimes-configurable","url":"https://github.com/openai/openai-go/releases/tag/v3.87.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":3,"composition":{"bugs":0,"features":2,"enhancements":0},"publishedAt":"2026-10-26T15:09:31.000Z"},{"id":"rel_qPJlshH54xYJRys1-X9Nf","title":"OpenAI Go SDK v3.91.0 adds agent context compaction and session review settings","path":"/release/rel_qPJlshH54xYJRys1-X9Nf-agent-context-compaction-settings-websocket-refusal-deltas-fixed","url":"https://github.com/openai/openai-go/releases/tag/v3.91.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":2,"composition":{"bugs":1,"features":8,"enhancements":0},"publishedAt":"2026-10-31T00:28:06.000Z"},{"id":"rel_Z6b__RNr3Jum2zYYFG9yO","title":"OpenAI Go SDK v3.90.0 adds usage tracking for image and web search in responses","path":"/release/rel_Z6b__RNr3Jum2zYYFG9yO-responses-track-image-and-web-search-usage-external-key-management-added","url":"https://github.com/openai/openai-go/releases/tag/v3.90.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":2,"composition":{"bugs":1,"features":3,"enhancements":0},"publishedAt":"2026-10-28T21:33:23.000Z"},{"id":"rel_y56Y3OF-1EIPx47aYO2zZ","title":"OpenAI Python SDK v3.40.0 adds prompt cache breakpoint TTL options","path":"/release/rel_y56Y3OF-1EIPx47aYO2zZ-prompt-cache-breakpoints-get-ttl-options","url":"https://github.com/openai/openai-python/releases/tag/v3.40.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":2,"composition":{"bugs":1,"features":1,"enhancements":0},"publishedAt":"2026-10-27T15:37:52.000Z"},{"id":"rel_ql4SVN9N0e2-k-oT5ESUb","title":"OpenAI Go SDK v3.89.0 adds prompt cache breakpoint TTL options","path":"/release/rel_ql4SVN9N0e2-k-oT5ESUb-prompt-cache-breakpoint-ttl-options-added-null-collections-isolated","url":"https://github.com/openai/openai-go/releases/tag/v3.89.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":2,"composition":{"bugs":2,"features":1,"enhancements":0},"publishedAt":"2026-10-27T15:38:32.000Z"},{"id":"rel_RJ9iARUNFADPfI6K0_mAh","title":"OpenAI Node SDK v7.45.1 clarifies prompt cache breakpoint TTL and adds google-auth-library v11 support","path":"/release/rel_RJ9iARUNFADPfI6K0_mAh-prompt-cache-breakpoint-ttl-clarified-vertex-supports-google-auth-library-v11","url":"https://github.com/openai/openai-node/releases/tag/v7.45.1","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":1,"composition":{"bugs":2,"features":0,"enhancements":0},"publishedAt":"2026-10-27T15:37:54.000Z"},{"id":"rel_Cz1ewjydlKZQlfqzTLNcv","title":"OpenAI Node SDK v7.45.0 adds gpt-6.1-terra model identifier","path":"/release/rel_Cz1ewjydlKZQlfqzTLNcv-gpt-6-1-terra-model-identifier-added","url":"https://github.com/openai/openai-node/releases/tag/v7.45.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":2,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-26T19:14:21.000Z"},{"id":"rel_bLz-dZtlFKjskB2U-NTAU","title":"OpenAI Python SDK v3.39.0 adds gpt-6.1-terra model identifier","path":"/release/rel_bLz-dZtlFKjskB2U-NTAU-gpt-6-1-terra-model-identifier-available","url":"https://github.com/openai/openai-python/releases/tag/v3.39.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":3,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-26T19:14:00.000Z"},{"id":"rel_nENFj8Nrv_IbFIK82tUmU","title":"OpenAI Go SDK v3.88.0 adds gpt-6.1-terra model identifier","path":"/release/rel_nENFj8Nrv_IbFIK82tUmU-gpt-6-1-terra-model-identifier-added","url":"https://github.com/openai/openai-go/releases/tag/v3.88.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":2,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-26T19:27:01.000Z"},{"id":"rel_WmstzSLoL9UObU26VyKYj","title":"OpenAI Go SDK v3.91.1 adds Live WebSocket guide","path":"/release/rel_WmstzSLoL9UObU26VyKYj-live-websocket-guide-added-to-docs","url":"https://github.com/openai/openai-go/releases/tag/v3.91.1","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":1,"composition":{"bugs":0,"features":0,"enhancements":1},"publishedAt":"2026-11-01T19:42:15.000Z"},{"id":"rel_D0U8tKGAXFWX6y5qOP4nu","title":"OpenAI Python SDK v3.42.1 preserves Responses WebSocket behavior and validates realtime auth","path":"/release/rel_D0U8tKGAXFWX6y5qOP4nu-responses-websocket-behavior-preserved-realtime-auth-profiles-validated","url":"https://github.com/openai/openai-python/releases/tag/v3.42.1","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":1,"composition":{"bugs":2,"features":0,"enhancements":1},"publishedAt":"2026-11-01T20:56:19.000Z"},{"id":"rel_ylZHV7lEyapGlaXXkrk29","title":"Codex CLI 0.165.0 adds Vertex AI provider and GPT-6.1 Terra","path":"/release/rel_ylZHV7lEyapGlaXXkrk29-vertex-ai-provider-added-gpt-6-1-terra-in-model-catalogs","url":"https://developers.openai.com/codex/changelog/#github-release-425927431","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"codex","name":"Codex"},"importance":4,"composition":{"bugs":6,"features":6,"enhancements":0},"publishedAt":"2026-10-27T00:00:00.000Z"},{"id":"rel_8BkJacLBZaovUaPmsX-qp","title":"Codex CLI 0.166.0 adds GPT-6.1 Terra and keyboard-driven transcript selection","path":"/release/rel_8BkJacLBZaovUaPmsX-qp-gpt-6-1-terra-lands-in-model-catalogs-transcript-selection-gets-keyboard-control","url":"https://developers.openai.com/codex/changelog/#github-release-428320524","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"codex","name":"Codex"},"importance":3,"composition":{"bugs":7,"features":6,"enhancements":0},"publishedAt":"2026-10-29T00:00:00.000Z"},{"id":"rel_GVnwz7Q4leUVNJVfqLNdq","title":"Claude Code v2.1.314 adds write-block setting and fixes turn after turn effort errors","path":"/release/rel_GVnwz7Q4leUVNJVfqLNdq-writes-outside-working-directories-now-blockable-effort-cap-error-fixed","url":"https://code.claude.com/docs/en/changelog#2-1-314","org":{"slug":"anthropic","name":"Anthropic","avatarUrl":"https://media.releases.sh/orgs/anthropic.png","githubHandle":"anthropics"},"product":{"slug":"claude-code","name":"Claude Code"},"importance":3,"composition":{"bugs":40,"features":4,"enhancements":1},"publishedAt":"2026-10-30T01:06:26.000Z"},{"id":"rel_mapQ4E72ZAef5Bh48WVjA","title":"Claude Code v2.1.312 stops Write from clobbering large CLAUDE.md files","path":"/release/rel_mapQ4E72ZAef5Bh48WVjA-write-no-longer-clobbers-large-claude-md-per-directory-skills-fixed","url":"https://code.claude.com/docs/en/changelog#2-1-312","org":{"slug":"anthropic","name":"Anthropic","avatarUrl":"https://media.releases.sh/orgs/anthropic.png","githubHandle":"anthropics"},"product":{"slug":"claude-code","name":"Claude Code"},"importance":3,"composition":{"bugs":18,"features":3,"enhancements":11},"publishedAt":"2026-10-27T20:24:20.000Z"},{"id":"rel_SUgZu2mcsNgA4M9QNDQ0a","title":"Claude Code v2.1.311 fixes startup freezes and 1M-context billing stops","path":"/release/rel_SUgZu2mcsNgA4M9QNDQ0a-long-claude-md-freezes-fixed-200k-sessions-now-compact-not-stop","url":"https://code.claude.com/docs/en/changelog#2-1-311","org":{"slug":"anthropic","name":"Anthropic","avatarUrl":"https://media.releases.sh/orgs/anthropic.png","githubHandle":"anthropics"},"product":{"slug":"claude-code","name":"Claude Code"},"importance":3,"composition":{"bugs":45,"features":2,"enhancements":6},"publishedAt":"2026-10-26T21:51:23.000Z"},{"id":"rel_RtkLb4jV3d50qor9FjLYd","title":"Claude Code v2.1.316 deprecates Opus 4.8 fast mode, adds subagent effort control","path":"/release/rel_RtkLb4jV3d50qor9FjLYd-opus-4-8-fast-mode-deprecated-subagent-effort-env-var-added","url":"https://code.claude.com/docs/en/changelog#2-1-316","org":{"slug":"anthropic","name":"Anthropic","avatarUrl":"https://media.releases.sh/orgs/anthropic.png","githubHandle":"anthropics"},"product":{"slug":"claude-code","name":"Claude Code"},"importance":4,"composition":{"bugs":24,"features":8,"enhancements":6},"publishedAt":"2026-10-31T01:01:35.000Z"},{"id":"rel_q7lKmiDVrUoX6VmnOilx_","title":"ChatGPT rolls out beta texting for your dot to Pro users in the US","path":"/release/rel_q7lKmiDVrUoX6VmnOilx_-text-your-dot-via-imessage-or-rcs-in-beta","url":"https://help.openai.com/en/articles/6825453-chatgpt-release-notes#texting-your-dot","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt","name":"ChatGPT"},"importance":4,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-28T00:00:00.000Z"},{"id":"rel_UwevlxtYDP84n9Htu4AmO","title":"ChatGPT rolls out account switching on iOS and Android","path":"/release/rel_UwevlxtYDP84n9Htu4AmO-chatgpt-apps-now-switch-between-2-signed-in-accounts","url":"https://help.openai.com/en/articles/6825453-chatgpt-release-notes#account-switching-on-ios-and-android","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt","name":"ChatGPT"},"importance":3,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-30T00:00:00.000Z"},{"id":"rel_DlR_k1dbz6AnMO2uX90Bk","title":"ChatGPT connects iCloud Mail and Calendar for email and schedule lookups","path":"/release/rel_DlR_k1dbz6AnMO2uX90Bk-icloud-mail-and-calendar-now-connect-to-chatgpt","url":"https://help.openai.com/en/articles/6825453-chatgpt-release-notes#icloud-mail-and-calendar-plugins-in-chatgpt","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt","name":"ChatGPT"},"importance":3,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-27T00:00:00.000Z"},{"id":"rel_BIKlM7XFvprZgI3HgREmF","title":"OpenAI makes GPT-6 in ChatGPT more concise and less flattering for Free and Go users","path":"/release/rel_BIKlM7XFvprZgI3HgREmF-gpt-6-in-chat-more-concise-less-flattering-plus-pro-fallback-only","url":"https://help.openai.com/en/articles/9624314-model-release-notes#gpt-6-in-chatgpt-update","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"models","name":"Models"},"importance":3,"composition":{"bugs":0,"features":0,"enhancements":2},"publishedAt":"2026-10-29T00:00:00.000Z"},{"id":"rel_kNS1ao-uEz3qakFgQJch-","title":"ChatGPT Enterprise & Edu retires automatic Instant-to-thinking model switching","path":"/release/rel_kNS1ao-uEz3qakFgQJch--automatic-instant-to-thinking-switching-retired","url":"https://help.openai.com/en/articles/10128477-chatgpt-enterprise-edu-release-notes#changes-to-automatic-thinking","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt-enterprise-edu","name":"ChatGPT Enterprise & Edu"},"importance":4,"composition":{"bugs":0,"features":0,"enhancements":1},"publishedAt":"2026-10-28T00:00:00.000Z"},{"id":"rel_xlxjl8JvcdJMwuZqP4uYq","title":"ChatGPT doubles Work and Codex usage on Pro and Business Premium plans","path":"/release/rel_xlxjl8JvcdJMwuZqP4uYq-pro-and-business-premium-work-and-codex-allowances-double","url":"https://help.openai.com/en/articles/6825453-chatgpt-release-notes#more-work-and-codex-usage-on-pro-and-business-premium-plans","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt","name":"ChatGPT"},"importance":4,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-29T00:00:00.000Z"},{"id":"rel_yzaoey9-A32Gja1tXigmi","title":"OpenAI API offers ChatGPT Pro and Business subscribers up to $500 monthly in free credits","path":"/release/rel_yzaoey9-A32Gja1tXigmi-chatgpt-pro-business-subscribers-get-up-to-500-mo-free-api-credits","url":"https://developers.openai.com/api/docs/changelog#2026-10-28","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"api","name":"API"},"importance":4,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-28T00:00:00.000Z"},{"id":"rel_JEHWGD44eDGlqN-vcRnVL","title":"ChatGPT Business adds promotional API credits for eligible workspaces","path":"/release/rel_JEHWGD44eDGlqN-vcRnVL-eligible-chatgpt-business-workspaces-can-activate-api-credits","url":"https://help.openai.com/en/articles/11391654-chatgpt-business-release-notes#api-credits-for-eligible-business-workspaces","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt-business","name":"ChatGPT Business"},"importance":3,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-28T00:00:00.000Z"},{"id":"rel_S9kt-lYrjrwjRvgDw42RV","title":"ChatGPT Pro and Business subscribers can activate promotional API credits","path":"/release/rel_S9kt-lYrjrwjRvgDw42RV-promotional-api-credits-offered-to-pro-business-plans","url":"https://help.openai.com/en/articles/6825453-chatgpt-release-notes#api-credits-for-eligible-chatgpt-pro-and-business-plans","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt","name":"ChatGPT"},"importance":3,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-28T00:00:00.000Z"},{"id":"rel_w5k4pmjpV1ne9RsUqdZe2","title":"Claude Platform launches Usage and Cost Analytics API in beta","path":"/release/rel_w5k4pmjpV1ne9RsUqdZe2-usage-and-cost-analytics-api-launches-in-beta","url":"https://platform.claude.com/docs/en/release-notes/overview#usage-and-cost-analytics-api-beta-and-legacy-usage-cost-endp","org":{"slug":"anthropic","name":"Anthropic","avatarUrl":"https://media.releases.sh/orgs/anthropic.png","githubHandle":"anthropics"},"product":{"slug":"claude-platform","name":"Claude Platform"},"importance":3,"composition":{"bugs":0,"features":1,"enhancements":1},"publishedAt":"2026-10-29T00:00:00.000Z"},{"id":"rel_cUdw9pwBkz_45EydzHTNr","title":"Claude Platform raises Messages API max request size to 128 MB","path":"/release/rel_cUdw9pwBkz_45EydzHTNr-max-request-size-now-128-mb-up-from-32-mb","url":"https://platform.claude.com/docs/en/release-notes/overview#maximum-request-size-raised-to-128-mb","org":{"slug":"anthropic","name":"Anthropic","avatarUrl":"https://media.releases.sh/orgs/anthropic.png","githubHandle":"anthropics"},"product":{"slug":"claude-platform","name":"Claude Platform"},"importance":3,"composition":{"bugs":0,"features":0,"enhancements":1},"publishedAt":"2026-10-28T00:00:00.000Z"},{"id":"rel_dRmjmW0FqWHjcIfzn0yRj","title":"OpenAI reorganizes Model Spec around five core values","path":"/release/rel_dRmjmW0FqWHjcIfzn0yRj-model-spec-restructured-around-five-core-values","url":"https://help.openai.com/en/articles/9624314-model-release-notes#updates-to-the-openai-model-spec-october-29-2026","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"models","name":"Models"},"importance":3,"composition":{"bugs":0,"features":0,"enhancements":5},"publishedAt":"2026-10-29T00:00:00.000Z"}],"sections":[{"heading":"GPT-6 goes to the bench","anchor":"gpt-6-goes-to-the-bench","lede":"The week's biggest story is OpenAI's decision that its frontier reasoning model needed a domain-specific variant.","releaseIds":["rel_38Gl2Tn-hz5ZWf5wvjjAG","rel_0U_gh2KVd7InoN_sa_E9Y","rel_klAXbgyARrcTjU4U3Y2mN","rel_9N2KJq9et8LGXygmZc9AR"],"releases":[{"id":"rel_38Gl2Tn-hz5ZWf5wvjjAG","title":"OpenAI launches GPT-6 Rosalind Research for multistep life sciences work","path":"/release/rel_38Gl2Tn-hz5ZWf5wvjjAG-gpt-6-rosalind-research-opens-to-approved-customers","url":"https://help.openai.com/en/articles/9624314-model-release-notes#gpt-6-rosalind-research","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"models","name":"Models"},"importance":5,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-30T00:00:00.000Z"},{"id":"rel_0U_gh2KVd7InoN_sa_E9Y","title":"ChatGPT Enterprise & Edu rolls out GPT-6 Rosalind Research in two-week Admin Preview","path":"/release/rel_0U_gh2KVd7InoN_sa_E9Y-gpt-6-rosalind-research-ships-in-off-by-default-admin-preview","url":"https://help.openai.com/en/articles/10128477-chatgpt-enterprise-edu-release-notes#gpt-6-rosalind-research","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt-enterprise-edu","name":"ChatGPT Enterprise & Edu"},"importance":4,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-30T00:00:00.000Z"},{"id":"rel_klAXbgyARrcTjU4U3Y2mN","title":"ChatGPT Business rolls out GPT-6 Rosalind Research to approved life sciences customers","path":"/release/rel_klAXbgyARrcTjU4U3Y2mN-gpt-6-rosalind-research-limited-rollout-for-life-sciences","url":"https://help.openai.com/en/articles/11391654-chatgpt-business-release-notes#gpt-6-rosalind-research","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt-business","name":"ChatGPT Business"},"importance":4,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-30T00:00:00.000Z"},{"id":"rel_9N2KJq9et8LGXygmZc9AR","title":"Anthropic reports early progress on accelerating drug development with Claude","path":"/release/rel_9N2KJq9et8LGXygmZc9AR-early-drug-development-progress-with-claude","url":"https://www.anthropic.com/research/early-progress-drug-development","org":{"slug":"anthropic","name":"Anthropic","avatarUrl":"https://media.releases.sh/orgs/anthropic.png","githubHandle":"anthropics"},"product":null,"importance":3,"composition":null,"publishedAt":"2026-10-27T00:00:00.000Z"}]},{"heading":"Agents become the unit of work in the SDKs","anchor":"agents-become-the-unit-of-work-in-the-sdks","lede":"The OpenAI SDKs converged on the same set of primitives across languages this week.","releaseIds":["rel_wQr5aYJ1ee0uoaaGBVDLs","rel_CS8H8QNz1c4vqpSwgC82Z","rel_HVVeaMYyBgZScWsPV_o7H","rel_VnaN3lCvKUoYe6HvI-aKF","rel_8_kg7w__zWWj4yPt_Znbu","rel_T52RUsDfAhuNiI4L6brYc","rel_l0xTEcPaRgui6oYen6wn8","rel_qPJlshH54xYJRys1-X9Nf","rel_Z6b__RNr3Jum2zYYFG9yO","rel_y56Y3OF-1EIPx47aYO2zZ","rel_ql4SVN9N0e2-k-oT5ESUb","rel_RJ9iARUNFADPfI6K0_mAh","rel_Cz1ewjydlKZQlfqzTLNcv","rel_bLz-dZtlFKjskB2U-NTAU","rel_nENFj8Nrv_IbFIK82tUmU","rel_WmstzSLoL9UObU26VyKYj","rel_D0U8tKGAXFWX6y5qOP4nu"],"releases":[{"id":"rel_wQr5aYJ1ee0uoaaGBVDLs","title":"OpenAI Node SDK v7.44.0 adds agent automations and delegation controls","path":"/release/rel_wQr5aYJ1ee0uoaaGBVDLs-agent-automations-api-added-live-decision-item-types-break","url":"https://github.com/openai/openai-node/releases/tag/v7.44.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":4,"composition":{"bugs":0,"features":3,"enhancements":1},"publishedAt":"2026-10-26T15:31:46.000Z"},{"id":"rel_CS8H8QNz1c4vqpSwgC82Z","title":"OpenAI Node SDK v7.46.0 adds agent session review and tool settings","path":"/release/rel_CS8H8QNz1c4vqpSwgC82Z-agent-session-review-and-tool-settings-external-key-management-added","url":"https://github.com/openai/openai-node/releases/tag/v7.46.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":3,"composition":{"bugs":1,"features":9,"enhancements":0},"publishedAt":"2026-10-31T00:27:06.000Z"},{"id":"rel_HVVeaMYyBgZScWsPV_o7H","title":"OpenAI Node SDK v7.47.0 adds typed Responses WebSocket warmup","path":"/release/rel_HVVeaMYyBgZScWsPV_o7H-responses-websocket-warmup-now-typed","url":"https://github.com/openai/openai-node/releases/tag/v7.47.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":2,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-31T04:46:03.000Z"},{"id":"rel_VnaN3lCvKUoYe6HvI-aKF","title":"OpenAI Python SDK v3.38.0 adds delegation controls and prompt cache lifetimes","path":"/release/rel_VnaN3lCvKUoYe6HvI-aKF-delegation-controls-and-prompt-cache-lifetimes-added","url":"https://github.com/openai/openai-python/releases/tag/v3.38.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":3,"composition":{"bugs":5,"features":2,"enhancements":0},"publishedAt":"2026-10-26T16:35:03.000Z"},{"id":"rel_8_kg7w__zWWj4yPt_Znbu","title":"OpenAI Python SDK v3.41.0 adds response tool usage models","path":"/release/rel_8_kg7w__zWWj4yPt_Znbu-response-tool-usage-models-agent-sessions-filter-by-metadata","url":"https://github.com/openai/openai-python/releases/tag/v3.41.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":3,"composition":{"bugs":3,"features":4,"enhancements":0},"publishedAt":"2026-10-28T22:26:54.000Z"},{"id":"rel_T52RUsDfAhuNiI4L6brYc","title":"OpenAI Python SDK v3.42.0 adds agent context compaction and session review settings","path":"/release/rel_T52RUsDfAhuNiI4L6brYc-agent-context-compaction-and-session-review-settings-added","url":"https://github.com/openai/openai-python/releases/tag/v3.42.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":3,"composition":{"bugs":1,"features":9,"enhancements":0},"publishedAt":"2026-10-31T00:26:41.000Z"},{"id":"rel_l0xTEcPaRgui6oYen6wn8","title":"OpenAI Go SDK v3.87.0 adds delegation controls and prompt cache lifetimes","path":"/release/rel_l0xTEcPaRgui6oYen6wn8-delegation-controls-prompt-cache-lifetimes-configurable","url":"https://github.com/openai/openai-go/releases/tag/v3.87.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":3,"composition":{"bugs":0,"features":2,"enhancements":0},"publishedAt":"2026-10-26T15:09:31.000Z"},{"id":"rel_qPJlshH54xYJRys1-X9Nf","title":"OpenAI Go SDK v3.91.0 adds agent context compaction and session review settings","path":"/release/rel_qPJlshH54xYJRys1-X9Nf-agent-context-compaction-settings-websocket-refusal-deltas-fixed","url":"https://github.com/openai/openai-go/releases/tag/v3.91.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":2,"composition":{"bugs":1,"features":8,"enhancements":0},"publishedAt":"2026-10-31T00:28:06.000Z"},{"id":"rel_Z6b__RNr3Jum2zYYFG9yO","title":"OpenAI Go SDK v3.90.0 adds usage tracking for image and web search in responses","path":"/release/rel_Z6b__RNr3Jum2zYYFG9yO-responses-track-image-and-web-search-usage-external-key-management-added","url":"https://github.com/openai/openai-go/releases/tag/v3.90.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":2,"composition":{"bugs":1,"features":3,"enhancements":0},"publishedAt":"2026-10-28T21:33:23.000Z"},{"id":"rel_y56Y3OF-1EIPx47aYO2zZ","title":"OpenAI Python SDK v3.40.0 adds prompt cache breakpoint TTL options","path":"/release/rel_y56Y3OF-1EIPx47aYO2zZ-prompt-cache-breakpoints-get-ttl-options","url":"https://github.com/openai/openai-python/releases/tag/v3.40.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":2,"composition":{"bugs":1,"features":1,"enhancements":0},"publishedAt":"2026-10-27T15:37:52.000Z"},{"id":"rel_ql4SVN9N0e2-k-oT5ESUb","title":"OpenAI Go SDK v3.89.0 adds prompt cache breakpoint TTL options","path":"/release/rel_ql4SVN9N0e2-k-oT5ESUb-prompt-cache-breakpoint-ttl-options-added-null-collections-isolated","url":"https://github.com/openai/openai-go/releases/tag/v3.89.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":2,"composition":{"bugs":2,"features":1,"enhancements":0},"publishedAt":"2026-10-27T15:38:32.000Z"},{"id":"rel_RJ9iARUNFADPfI6K0_mAh","title":"OpenAI Node SDK v7.45.1 clarifies prompt cache breakpoint TTL and adds google-auth-library v11 support","path":"/release/rel_RJ9iARUNFADPfI6K0_mAh-prompt-cache-breakpoint-ttl-clarified-vertex-supports-google-auth-library-v11","url":"https://github.com/openai/openai-node/releases/tag/v7.45.1","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":1,"composition":{"bugs":2,"features":0,"enhancements":0},"publishedAt":"2026-10-27T15:37:54.000Z"},{"id":"rel_Cz1ewjydlKZQlfqzTLNcv","title":"OpenAI Node SDK v7.45.0 adds gpt-6.1-terra model identifier","path":"/release/rel_Cz1ewjydlKZQlfqzTLNcv-gpt-6-1-terra-model-identifier-added","url":"https://github.com/openai/openai-node/releases/tag/v7.45.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":2,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-26T19:14:21.000Z"},{"id":"rel_bLz-dZtlFKjskB2U-NTAU","title":"OpenAI Python SDK v3.39.0 adds gpt-6.1-terra model identifier","path":"/release/rel_bLz-dZtlFKjskB2U-NTAU-gpt-6-1-terra-model-identifier-available","url":"https://github.com/openai/openai-python/releases/tag/v3.39.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":3,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-26T19:14:00.000Z"},{"id":"rel_nENFj8Nrv_IbFIK82tUmU","title":"OpenAI Go SDK v3.88.0 adds gpt-6.1-terra model identifier","path":"/release/rel_nENFj8Nrv_IbFIK82tUmU-gpt-6-1-terra-model-identifier-added","url":"https://github.com/openai/openai-go/releases/tag/v3.88.0","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":2,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-26T19:27:01.000Z"},{"id":"rel_WmstzSLoL9UObU26VyKYj","title":"OpenAI Go SDK v3.91.1 adds Live WebSocket guide","path":"/release/rel_WmstzSLoL9UObU26VyKYj-live-websocket-guide-added-to-docs","url":"https://github.com/openai/openai-go/releases/tag/v3.91.1","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":1,"composition":{"bugs":0,"features":0,"enhancements":1},"publishedAt":"2026-11-01T19:42:15.000Z"},{"id":"rel_D0U8tKGAXFWX6y5qOP4nu","title":"OpenAI Python SDK v3.42.1 preserves Responses WebSocket behavior and validates realtime auth","path":"/release/rel_D0U8tKGAXFWX6y5qOP4nu-responses-websocket-behavior-preserved-realtime-auth-profiles-validated","url":"https://github.com/openai/openai-python/releases/tag/v3.42.1","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":null,"importance":1,"composition":{"bugs":2,"features":0,"enhancements":1},"publishedAt":"2026-11-01T20:56:19.000Z"}]},{"heading":"Codex and Claude Code harden the local agent loop","anchor":"codex-and-claude-code-harden-the-local-agent-loop","lede":"Codex CLI added a built-in Vertex AI provider and GPT-6.1 Terra in the model catalogs, turned on memory v2 and mid-conversation reasoning-effort changes by default, and closed a marketplace URL validation gap that could send Git to a different host than the one checked.","releaseIds":["rel_ylZHV7lEyapGlaXXkrk29","rel_8BkJacLBZaovUaPmsX-qp","rel_GVnwz7Q4leUVNJVfqLNdq","rel_mapQ4E72ZAef5Bh48WVjA","rel_SUgZu2mcsNgA4M9QNDQ0a","rel_RtkLb4jV3d50qor9FjLYd"],"releases":[{"id":"rel_ylZHV7lEyapGlaXXkrk29","title":"Codex CLI 0.165.0 adds Vertex AI provider and GPT-6.1 Terra","path":"/release/rel_ylZHV7lEyapGlaXXkrk29-vertex-ai-provider-added-gpt-6-1-terra-in-model-catalogs","url":"https://developers.openai.com/codex/changelog/#github-release-425927431","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"codex","name":"Codex"},"importance":4,"composition":{"bugs":6,"features":6,"enhancements":0},"publishedAt":"2026-10-27T00:00:00.000Z"},{"id":"rel_8BkJacLBZaovUaPmsX-qp","title":"Codex CLI 0.166.0 adds GPT-6.1 Terra and keyboard-driven transcript selection","path":"/release/rel_8BkJacLBZaovUaPmsX-qp-gpt-6-1-terra-lands-in-model-catalogs-transcript-selection-gets-keyboard-control","url":"https://developers.openai.com/codex/changelog/#github-release-428320524","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"codex","name":"Codex"},"importance":3,"composition":{"bugs":7,"features":6,"enhancements":0},"publishedAt":"2026-10-29T00:00:00.000Z"},{"id":"rel_GVnwz7Q4leUVNJVfqLNdq","title":"Claude Code v2.1.314 adds write-block setting and fixes turn after turn effort errors","path":"/release/rel_GVnwz7Q4leUVNJVfqLNdq-writes-outside-working-directories-now-blockable-effort-cap-error-fixed","url":"https://code.claude.com/docs/en/changelog#2-1-314","org":{"slug":"anthropic","name":"Anthropic","avatarUrl":"https://media.releases.sh/orgs/anthropic.png","githubHandle":"anthropics"},"product":{"slug":"claude-code","name":"Claude Code"},"importance":3,"composition":{"bugs":40,"features":4,"enhancements":1},"publishedAt":"2026-10-30T01:06:26.000Z"},{"id":"rel_mapQ4E72ZAef5Bh48WVjA","title":"Claude Code v2.1.312 stops Write from clobbering large CLAUDE.md files","path":"/release/rel_mapQ4E72ZAef5Bh48WVjA-write-no-longer-clobbers-large-claude-md-per-directory-skills-fixed","url":"https://code.claude.com/docs/en/changelog#2-1-312","org":{"slug":"anthropic","name":"Anthropic","avatarUrl":"https://media.releases.sh/orgs/anthropic.png","githubHandle":"anthropics"},"product":{"slug":"claude-code","name":"Claude Code"},"importance":3,"composition":{"bugs":18,"features":3,"enhancements":11},"publishedAt":"2026-10-27T20:24:20.000Z"},{"id":"rel_SUgZu2mcsNgA4M9QNDQ0a","title":"Claude Code v2.1.311 fixes startup freezes and 1M-context billing stops","path":"/release/rel_SUgZu2mcsNgA4M9QNDQ0a-long-claude-md-freezes-fixed-200k-sessions-now-compact-not-stop","url":"https://code.claude.com/docs/en/changelog#2-1-311","org":{"slug":"anthropic","name":"Anthropic","avatarUrl":"https://media.releases.sh/orgs/anthropic.png","githubHandle":"anthropics"},"product":{"slug":"claude-code","name":"Claude Code"},"importance":3,"composition":{"bugs":45,"features":2,"enhancements":6},"publishedAt":"2026-10-26T21:51:23.000Z"},{"id":"rel_RtkLb4jV3d50qor9FjLYd","title":"Claude Code v2.1.316 deprecates Opus 4.8 fast mode, adds subagent effort control","path":"/release/rel_RtkLb4jV3d50qor9FjLYd-opus-4-8-fast-mode-deprecated-subagent-effort-env-var-added","url":"https://code.claude.com/docs/en/changelog#2-1-316","org":{"slug":"anthropic","name":"Anthropic","avatarUrl":"https://media.releases.sh/orgs/anthropic.png","githubHandle":"anthropics"},"product":{"slug":"claude-code","name":"Claude Code"},"importance":4,"composition":{"bugs":24,"features":8,"enhancements":6},"publishedAt":"2026-10-31T01:01:35.000Z"}]},{"heading":"Product surface and platform plumbing","anchor":"product-surface-and-platform-plumbing","lede":"ChatGPT itself got more conversational infrastructure than model news: texting your dot via iMessage or RCS began a U.S. beta for Pro users, account switching arrived on iOS and Android, iCloud Mail and Calendar became connectable apps, and Free and Go users got more concise, less flattering GPT-6 replies.","releaseIds":["rel_q7lKmiDVrUoX6VmnOilx_","rel_UwevlxtYDP84n9Htu4AmO","rel_DlR_k1dbz6AnMO2uX90Bk","rel_BIKlM7XFvprZgI3HgREmF","rel_kNS1ao-uEz3qakFgQJch-","rel_xlxjl8JvcdJMwuZqP4uYq","rel_yzaoey9-A32Gja1tXigmi","rel_JEHWGD44eDGlqN-vcRnVL","rel_S9kt-lYrjrwjRvgDw42RV","rel_w5k4pmjpV1ne9RsUqdZe2","rel_cUdw9pwBkz_45EydzHTNr","rel_dRmjmW0FqWHjcIfzn0yRj"],"releases":[{"id":"rel_q7lKmiDVrUoX6VmnOilx_","title":"ChatGPT rolls out beta texting for your dot to Pro users in the US","path":"/release/rel_q7lKmiDVrUoX6VmnOilx_-text-your-dot-via-imessage-or-rcs-in-beta","url":"https://help.openai.com/en/articles/6825453-chatgpt-release-notes#texting-your-dot","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt","name":"ChatGPT"},"importance":4,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-28T00:00:00.000Z"},{"id":"rel_UwevlxtYDP84n9Htu4AmO","title":"ChatGPT rolls out account switching on iOS and Android","path":"/release/rel_UwevlxtYDP84n9Htu4AmO-chatgpt-apps-now-switch-between-2-signed-in-accounts","url":"https://help.openai.com/en/articles/6825453-chatgpt-release-notes#account-switching-on-ios-and-android","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt","name":"ChatGPT"},"importance":3,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-30T00:00:00.000Z"},{"id":"rel_DlR_k1dbz6AnMO2uX90Bk","title":"ChatGPT connects iCloud Mail and Calendar for email and schedule lookups","path":"/release/rel_DlR_k1dbz6AnMO2uX90Bk-icloud-mail-and-calendar-now-connect-to-chatgpt","url":"https://help.openai.com/en/articles/6825453-chatgpt-release-notes#icloud-mail-and-calendar-plugins-in-chatgpt","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt","name":"ChatGPT"},"importance":3,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-27T00:00:00.000Z"},{"id":"rel_BIKlM7XFvprZgI3HgREmF","title":"OpenAI makes GPT-6 in ChatGPT more concise and less flattering for Free and Go users","path":"/release/rel_BIKlM7XFvprZgI3HgREmF-gpt-6-in-chat-more-concise-less-flattering-plus-pro-fallback-only","url":"https://help.openai.com/en/articles/9624314-model-release-notes#gpt-6-in-chatgpt-update","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"models","name":"Models"},"importance":3,"composition":{"bugs":0,"features":0,"enhancements":2},"publishedAt":"2026-10-29T00:00:00.000Z"},{"id":"rel_kNS1ao-uEz3qakFgQJch-","title":"ChatGPT Enterprise & Edu retires automatic Instant-to-thinking model switching","path":"/release/rel_kNS1ao-uEz3qakFgQJch--automatic-instant-to-thinking-switching-retired","url":"https://help.openai.com/en/articles/10128477-chatgpt-enterprise-edu-release-notes#changes-to-automatic-thinking","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt-enterprise-edu","name":"ChatGPT Enterprise & Edu"},"importance":4,"composition":{"bugs":0,"features":0,"enhancements":1},"publishedAt":"2026-10-28T00:00:00.000Z"},{"id":"rel_xlxjl8JvcdJMwuZqP4uYq","title":"ChatGPT doubles Work and Codex usage on Pro and Business Premium plans","path":"/release/rel_xlxjl8JvcdJMwuZqP4uYq-pro-and-business-premium-work-and-codex-allowances-double","url":"https://help.openai.com/en/articles/6825453-chatgpt-release-notes#more-work-and-codex-usage-on-pro-and-business-premium-plans","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt","name":"ChatGPT"},"importance":4,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-29T00:00:00.000Z"},{"id":"rel_yzaoey9-A32Gja1tXigmi","title":"OpenAI API offers ChatGPT Pro and Business subscribers up to $500 monthly in free credits","path":"/release/rel_yzaoey9-A32Gja1tXigmi-chatgpt-pro-business-subscribers-get-up-to-500-mo-free-api-credits","url":"https://developers.openai.com/api/docs/changelog#2026-10-28","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"api","name":"API"},"importance":4,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-28T00:00:00.000Z"},{"id":"rel_JEHWGD44eDGlqN-vcRnVL","title":"ChatGPT Business adds promotional API credits for eligible workspaces","path":"/release/rel_JEHWGD44eDGlqN-vcRnVL-eligible-chatgpt-business-workspaces-can-activate-api-credits","url":"https://help.openai.com/en/articles/11391654-chatgpt-business-release-notes#api-credits-for-eligible-business-workspaces","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt-business","name":"ChatGPT Business"},"importance":3,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-28T00:00:00.000Z"},{"id":"rel_S9kt-lYrjrwjRvgDw42RV","title":"ChatGPT Pro and Business subscribers can activate promotional API credits","path":"/release/rel_S9kt-lYrjrwjRvgDw42RV-promotional-api-credits-offered-to-pro-business-plans","url":"https://help.openai.com/en/articles/6825453-chatgpt-release-notes#api-credits-for-eligible-chatgpt-pro-and-business-plans","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"chatgpt","name":"ChatGPT"},"importance":3,"composition":{"bugs":0,"features":1,"enhancements":0},"publishedAt":"2026-10-28T00:00:00.000Z"},{"id":"rel_w5k4pmjpV1ne9RsUqdZe2","title":"Claude Platform launches Usage and Cost Analytics API in beta","path":"/release/rel_w5k4pmjpV1ne9RsUqdZe2-usage-and-cost-analytics-api-launches-in-beta","url":"https://platform.claude.com/docs/en/release-notes/overview#usage-and-cost-analytics-api-beta-and-legacy-usage-cost-endp","org":{"slug":"anthropic","name":"Anthropic","avatarUrl":"https://media.releases.sh/orgs/anthropic.png","githubHandle":"anthropics"},"product":{"slug":"claude-platform","name":"Claude Platform"},"importance":3,"composition":{"bugs":0,"features":1,"enhancements":1},"publishedAt":"2026-10-29T00:00:00.000Z"},{"id":"rel_cUdw9pwBkz_45EydzHTNr","title":"Claude Platform raises Messages API max request size to 128 MB","path":"/release/rel_cUdw9pwBkz_45EydzHTNr-max-request-size-now-128-mb-up-from-32-mb","url":"https://platform.claude.com/docs/en/release-notes/overview#maximum-request-size-raised-to-128-mb","org":{"slug":"anthropic","name":"Anthropic","avatarUrl":"https://media.releases.sh/orgs/anthropic.png","githubHandle":"anthropics"},"product":{"slug":"claude-platform","name":"Claude Platform"},"importance":3,"composition":{"bugs":0,"features":0,"enhancements":1},"publishedAt":"2026-10-28T00:00:00.000Z"},{"id":"rel_dRmjmW0FqWHjcIfzn0yRj","title":"OpenAI reorganizes Model Spec around five core values","path":"/release/rel_dRmjmW0FqWHjcIfzn0yRj-model-spec-restructured-around-five-core-values","url":"https://help.openai.com/en/articles/9624314-model-release-notes#updates-to-the-openai-model-spec-october-29-2026","org":{"slug":"openai","name":"OpenAI","avatarUrl":"https://media.releases.sh/orgs/openai.png","githubHandle":"openai"},"product":{"slug":"models","name":"Models"},"importance":3,"composition":{"bugs":0,"features":0,"enhancements":5},"publishedAt":"2026-10-29T00:00:00.000Z"}]}]}}