renovate-bot opened a new pull request, #8951: URL: https://github.com/apache/texera/pull/8951
> ℹ️ **Note** > > This PR body was truncated due to platform limits. This PR contains the following updates: | Package | Change | [Age](https://docs.renovatebot.com/merge-confidence/) | [Confidence](https://docs.renovatebot.com/merge-confidence/) | |---|---|---|---| | [@ai-sdk/openai](https://ai-sdk.dev/docs) ([source](https://redirect.github.com/vercel/ai/tree/HEAD/packages/openai)) | [`4.0.36` → `4.0.91`](https://renovatebot.com/diffs/npm/@ai-sdk%2fopenai/4.0.36/4.0.91) |  |  | | [ai](https://ai-sdk.dev/docs) ([source](https://redirect.github.com/vercel/ai/tree/HEAD/packages/ai)) | [`7.0.58` → `7.0.137`](https://renovatebot.com/diffs/npm/ai/7.0.58/7.0.137) |  |  | | [elysia](https://redirect.github.com/elysiajs/elysia) | [`1.4.29` → `1.4.30`](https://renovatebot.com/diffs/npm/elysia/1.4.29/1.4.30) |  |  | | [prettier](https://prettier.io) ([source](https://redirect.github.com/prettier/prettier)) | [`3.9.6` → `3.9.10`](https://renovatebot.com/diffs/npm/prettier/3.9.6/3.9.10) |  |  | | [tsx](https://tsx.hirok.io) ([source](https://redirect.github.com/privatenumber/tsx)) | [`4.23.12` → `4.23.15`](https://renovatebot.com/diffs/npm/tsx/4.23.12/4.23.15) |  |  | --- ### Release Notes <details> <summary>vercel/ai (@​ai-sdk/openai)</summary> ### [`v4.0.91`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4091) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`5cf13d2`](https://redirect.github.com/vercel/ai/commit/5cf13d2): feat: map the top level embedding dimensions to provider settings - [`72e9d38`](https://redirect.github.com/vercel/ai/commit/72e9d38): feat(openai): send image detail in decisions - [`fb933f9`](https://redirect.github.com/vercel/ai/commit/fb933f9): Report failed and incomplete web searches as tool errors instead of successful tool results in OpenAI Responses API calls and batch results. - Updated dependencies \[[`4dbc902`](https://redirect.github.com/vercel/ai/commit/4dbc902)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.26 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.58 ### [`v4.0.90`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4090) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`6c59fb1`](https://redirect.github.com/vercel/ai/commit/6c59fb1): fix(openai): preserve sampling parameters for GPT-6 Sol and Luna when reasoning is disabled, accounting for reasoning effort updates in Responses requests ### [`v4.0.89`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4089) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`9184a35`](https://redirect.github.com/vercel/ai/commit/9184a35): Add ordered text, file, and JSON parts to experimental decision state. Support image input in OpenAI Decisions and language model adapters. Gateway retains its existing string and JSON request format and rejects files, as does TypeSafe AI. Arrays passed directly as state now contain decision state parts. Wrap JSON arrays in an object or a json part. Normalize all public state forms into an array of parts before calling decision providers. Providers receive text and JSON objects as text and json parts. Serialize decision file bytes as base64 in OpenTelemetry state attributes. Model-call spans record normalized state parts, while outer spans retain public state inputs. Preserve native JSON state in TypeSafe AI when state contains one JSON part. Label shared state in language-model decision prompts. - Updated dependencies \[[`9184a35`](https://redirect.github.com/vercel/ai/commit/9184a35)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.25 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.57 ### [`v4.0.88`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4088) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`32ff7ad`](https://redirect.github.com/vercel/ai/commit/32ff7ad): Preserve tool errors in OpenAI Chat Completions and Responses by wrapping their payloads in an `error` object. Send Google tool errors and denied executions under `functionResponse.response.error`. ### [`v4.0.87`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4087) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`bccc1b7`](https://redirect.github.com/vercel/ai/commit/bccc1b7): fix(openai): avoid duplicate references for stored messages with multiple text parts ### [`v4.0.86`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4086) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`d3bcad9`](https://redirect.github.com/vercel/ai/commit/d3bcad9): feat(provider): report decision refusals as refusal answers - Updated dependencies \[[`d3bcad9`](https://redirect.github.com/vercel/ai/commit/d3bcad9)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.24 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.56 ### [`v4.0.85`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4085) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`fc1e19e`](https://redirect.github.com/vercel/ai/commit/fc1e19e): fix: limit jsonl rows to 64mb during batch result downloads - [`99bdcd2`](https://redirect.github.com/vercel/ai/commit/99bdcd2): Replace the OpenAI structured-output wrapper for `decide` with the native Decisions API. Only `gpt-6-luna` is currently supported by this endpoint; other Responses models no longer work with `openai.decisionModel()` or its deprecated `evaluationModel()` alias. Responses API options, such as `reasoningEffort`, are no longer applied and produce unsupported-option warnings. The Decisions API supports `providerOptions.openai.safetyIdentifier`. Native token usage and confidence values are preserved, and per-question refusals fail the decision call. - [`268225c`](https://redirect.github.com/vercel/ai/commit/268225c): feat(provider): add a portable `max` reasoning level with native provider mappings, compatibility coercions, and budget-based fallback support. Compatibility notes: - **TypeScript source compatibility:** Adding `max` widens `LanguageModelV4CallOptions['reasoning']` and the derived `ReasoningLevel` type. Third-party providers and consumers with exhaustive switches or `Record<ReasoningLevel, ...>` mappings must handle `max` before upgrading. Ordinary calls using existing reasoning levels remain accepted. - **Amazon Bedrock:** Existing Nova 2 calls using portable `reasoning: 'xhigh'` now send `maxReasoningEffort: 'high'` instead of the unsupported `'max'`, and return a compatibility warning. Anthropic and OpenAI Bedrock effort mappings retain their existing `xhigh` behavior. Partial Bedrock reasoning configurations continue to preserve explicit values and derive missing fields; explicit effort/budget overrides and disabled thinking no longer emit warnings for portable mappings that are not sent. - **Fireworks:** Existing portable `minimal` and `xhigh` calls keep their request mappings to `low` and `high`, respectively, and now return compatibility warnings in generation and streaming results. Portable `max` maps to `high` with the same warning behavior. **Release classification:** This remains a patch changeset under the repository's explicit release policy, which uses patch releases for both fixes and features. The TypeScript source-compatibility caveat for exhaustive consumers is disclosed above. Maintainers can override this classification with a `major` label if they decide to align the provider-spec addition with a future AI SDK major release. - [`0af2f7c`](https://redirect.github.com/vercel/ai/commit/0af2f7c): chore: upgrade to TypeScript 6 - Updated dependencies \[[`fc1e19e`](https://redirect.github.com/vercel/ai/commit/fc1e19e)] - Updated dependencies \[[`5094ebd`](https://redirect.github.com/vercel/ai/commit/5094ebd)] - Updated dependencies \[[`268225c`](https://redirect.github.com/vercel/ai/commit/268225c)] - Updated dependencies \[[`0af2f7c`](https://redirect.github.com/vercel/ai/commit/0af2f7c)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.55 - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.23 ### [`v4.0.84`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4084) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`2959d35`](https://redirect.github.com/vercel/ai/commit/2959d35): feat: rename `evaluate` to `decide` - [`5b8e63b`](https://redirect.github.com/vercel/ai/commit/5b8e63b): feat(openai): warn when reasoningSummary is used with chat models - Updated dependencies \[[`2959d35`](https://redirect.github.com/vercel/ai/commit/2959d35)] - Updated dependencies \[[`59116e6`](https://redirect.github.com/vercel/ai/commit/59116e6)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.22 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.54 ### [`v4.0.83`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4083) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`8c65988`](https://redirect.github.com/vercel/ai/commit/8c65988): feat(ai): add telemetry to speech generation and transcription, including provider usage propagation and experimental streaming lifecycle callbacks - Updated dependencies \[[`8c65988`](https://redirect.github.com/vercel/ai/commit/8c65988)] - Updated dependencies \[[`527a163`](https://redirect.github.com/vercel/ai/commit/527a163)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.21 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.53 ### [`v4.0.82`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4082) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`040033b`](https://redirect.github.com/vercel/ai/commit/040033b): feat(openai): add GPT-6.1 Sol model support - [`ede5b89`](https://redirect.github.com/vercel/ai/commit/ede5b89): chore: migrate package builds from tsup to tsdown - Updated dependencies \[[`ede5b89`](https://redirect.github.com/vercel/ai/commit/ede5b89)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.20 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.52 ### [`v4.0.81`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4081) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`4e94782`](https://redirect.github.com/vercel/ai/commit/4e94782): fix(openai): rewrite recursive Zod schemas that use `allOf` wrappers ### [`v4.0.80`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4080) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`c2511c1`](https://redirect.github.com/vercel/ai/commit/c2511c1): fix: use standards-compliant User-Agent header - Updated dependencies \[[`c2511c1`](https://redirect.github.com/vercel/ai/commit/c2511c1)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.51 ### [`v4.0.79`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4079) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`e3605f6`](https://redirect.github.com/vercel/ai/commit/e3605f6): Fix streamed tool calls with missing, blank, or repeated IDs. - [`525efc5`](https://redirect.github.com/vercel/ai/commit/525efc5): feat(provider): advertise image model file and mask input support Use confirmed model IDs for capability declarations so unrecognized model names remain unknown. Include Together AI FLUX.2 Pro and Flex single-image editing, and allow asynchronous capability lookups and middleware overrides to resolve to unknown. Advertise QuiverAI Arrow 2 and Arrow 2 Telos file-input support, and mark Together AI Gemini image inputs unsupported by the current single-image request mapping. - Updated dependencies \[[`e3605f6`](https://redirect.github.com/vercel/ai/commit/e3605f6)] - Updated dependencies \[[`525efc5`](https://redirect.github.com/vercel/ai/commit/525efc5)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.50 - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.19 ### [`v4.0.78`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4078) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`94d5d6d`](https://redirect.github.com/vercel/ai/commit/94d5d6d): Support `reasoningEffortUpdate: 'none'` for GPT-6 Sol and Luna in request-level options and positioned system messages. Validate update efforts against the model's supported efforts, warning and omitting unsupported request-level updates and rejecting unsupported historical updates. ### [`v4.0.77`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4077) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`2abd503`](https://redirect.github.com/vercel/ai/commit/2abd503): Default OpenAI Responses function tools to `strict: false` when tool strict mode is omitted. - Updated dependencies \[[`af9597b`](https://redirect.github.com/vercel/ai/commit/af9597b)] - Updated dependencies \[[`bc49f78`](https://redirect.github.com/vercel/ai/commit/bc49f78)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.49 ### [`v4.0.75`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4075) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`ca31b89`](https://redirect.github.com/vercel/ai/commit/ca31b89): The OpenAI Responses provider now accepts `providerOptions.openai.reasoningEffortUpdate` on empty system messages and sends each update at its position in the conversation. This lets applications change reasoning effort during a conversation while preserving the prompt prefix for caching. ### [`v4.0.74`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4074) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`4cf5a99`](https://redirect.github.com/vercel/ai/commit/4cf5a99): fix(openai): classify Responses stream errors with null codes - [`771e74b`](https://redirect.github.com/vercel/ai/commit/771e74b): chore: enable dead code lint rules - Updated dependencies \[[`fe07867`](https://redirect.github.com/vercel/ai/commit/fe07867)] - Updated dependencies \[[`a4b0940`](https://redirect.github.com/vercel/ai/commit/a4b0940)] - Updated dependencies \[[`771e74b`](https://redirect.github.com/vercel/ai/commit/771e74b)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.47 ### [`v4.0.73`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4073) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`6d1f881`](https://redirect.github.com/vercel/ai/commit/6d1f881): fix(openai): expose Chat Completions audio transcripts as generated text - [`618dc11`](https://redirect.github.com/vercel/ai/commit/618dc11): feat: GPT-6 Sol and Luna model IDs - Updated dependencies \[[`ffb0e76`](https://redirect.github.com/vercel/ai/commit/ffb0e76)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.18 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.46 ### [`v4.0.72`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4072) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`411b3f2`](https://redirect.github.com/vercel/ai/commit/411b3f2): fix(openai): strip unsupported regex patterns from JSON schemas - [`e13c32f`](https://redirect.github.com/vercel/ai/commit/e13c32f): fix(openai): include speech provider options in requests ### [`v4.0.71`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4071) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`e76a0a3`](https://redirect.github.com/vercel/ai/commit/e76a0a3): fix(openai): cancel image edit URL downloads when the request is aborted - Updated dependencies \[[`2973485`](https://redirect.github.com/vercel/ai/commit/2973485)] - Updated dependencies \[[`a4db5ea`](https://redirect.github.com/vercel/ai/commit/a4db5ea)] - Updated dependencies \[[`2937ea2`](https://redirect.github.com/vercel/ai/commit/2937ea2)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.45 ### [`v4.0.70`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4070) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`fd75cee`](https://redirect.github.com/vercel/ai/commit/fd75cee): fix(openai): preserve provider file references in Responses tool results - [`1f5bb62`](https://redirect.github.com/vercel/ai/commit/1f5bb62): fix(openai): send assistant text as Responses easy input messages - Updated dependencies \[[`0455398`](https://redirect.github.com/vercel/ai/commit/0455398)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.44 ### [`v4.0.69`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4069) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`d4d96bf`](https://redirect.github.com/vercel/ai/commit/d4d96bf): Add experimental Choice, Score, and Boolean evaluations through `openai.evaluationModel()` and a shared structured language-model evaluation adapter in `@ai-sdk/provider-utils/experimental-evaluation`. Preserve exact labels and metadata, validate score bounds, and return prompted Boolean P(true) estimates validated to be in \[0, 1]. Boolean estimates are not guaranteed to be calibrated; application code chooses thresholds. - Updated dependencies \[[`215b25e`](https://redirect.github.com/vercel/ai/commit/215b25e)] - Updated dependencies \[[`d4d96bf`](https://redirect.github.com/vercel/ai/commit/d4d96bf)] - Updated dependencies \[[`a7dd893`](https://redirect.github.com/vercel/ai/commit/a7dd893)] - Updated dependencies \[[`3456e2c`](https://redirect.github.com/vercel/ai/commit/3456e2c)] - Updated dependencies \[[`c4e76de`](https://redirect.github.com/vercel/ai/commit/c4e76de)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.43 - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.17 ### [`v4.0.68`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4068) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - Updated dependencies \[[`91c2128`](https://redirect.github.com/vercel/ai/commit/91c2128)] - Updated dependencies \[[`2cd80b3`](https://redirect.github.com/vercel/ai/commit/2cd80b3)] - Updated dependencies \[[`d06bb2a`](https://redirect.github.com/vercel/ai/commit/d06bb2a)] - Updated dependencies \[[`123d71f`](https://redirect.github.com/vercel/ai/commit/123d71f)] - Updated dependencies \[[`2fa5e0e`](https://redirect.github.com/vercel/ai/commit/2fa5e0e)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.42 - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.16 ### [`v4.0.67`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4067) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`5c0054d`](https://redirect.github.com/vercel/ai/commit/5c0054d): Add optional browser-direct WebRTC for experimental client-delegated Live conversations alongside the existing WebSocket path. Exchange SDP through an application endpoint with `api.session`, configure server-owned data-channel permissions, and preserve committed React session ownership. Capture follows the selected sender track, borrowed tracks remain caller-owned, and disconnect recovery and finalization stay bounded. Applications continue to handle client delegation and submit context; Live session updates and Responses delegation remain unsupported. Serialize microphone sender changes and close the peer if detachment fails, without stopping borrowed tracks. Validate nonempty SDP setup answers with a bounded response body, and document the same-origin broker authentication contract. - [`39535af`](https://redirect.github.com/vercel/ai/commit/39535af): Add experimental OpenAI Live provider support through the unified `openai.experimental_realtime` factory for server WebSocket sessions with client delegation. Applications own their agents and tools, receive continuous audio/transcript events and delegation metadata, and return context through validated channels. Support immutable startup options, microphone mute controls, graceful session close, and cumulative voice usage. Responses delegation and Live session updates reject before sending. Route known Live model IDs to Live, allow an OpenAI-specific `api` override for early-access models, and preserve legacy Realtime defaults for unknown IDs. Token minting follows the same selection rules and rejects Live before requesting unsupported credentials. Extend the realtime v4 specification with optional server WebSocket configuration, per-connection raw-event parsers, and model-wide startup/finalization capabilities. This provider layer supplies connection settings and protocol mapping for server adapters; browser lifecycle and UI integration belong to the core runtime and framework hooks. Preserve Realtime client event IDs for session updates and audio appends, and correlate server errors with the originating client event. - Updated dependencies \[[`5c0054d`](https://redirect.github.com/vercel/ai/commit/5c0054d)] - Updated dependencies \[[`39535af`](https://redirect.github.com/vercel/ai/commit/39535af)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.15 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.41 ### [`v4.0.66`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4066) ##### Patch Changes - [`5ec21a6`](https://redirect.github.com/vercel/ai/commit/5ec21a6): fix: reject unsupported batch request types - [`7469a3b`](https://redirect.github.com/vercel/ai/commit/7469a3b): feat: support image generation requests in batches - [`0de8886`](https://redirect.github.com/vercel/ai/commit/0de8886): fix(openai): allow providers to disable web search source includes - [`d5e3024`](https://redirect.github.com/vercel/ai/commit/d5e3024): fix(openai): remove propertyNames from JSON Schema - Updated dependencies \[[`5ec21a6`](https://redirect.github.com/vercel/ai/commit/5ec21a6)] - Updated dependencies \[[`7469a3b`](https://redirect.github.com/vercel/ai/commit/7469a3b)] - Updated dependencies \[[`813bb36`](https://redirect.github.com/vercel/ai/commit/813bb36)] - Updated dependencies \[[`c43e4b7`](https://redirect.github.com/vercel/ai/commit/c43e4b7)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.14 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.40 ### [`v4.0.65`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4065) ##### Patch Changes - [`9942196`](https://redirect.github.com/vercel/ai/commit/9942196): feat(openai): add batch cancellation and listing - Updated dependencies \[[`9942196`](https://redirect.github.com/vercel/ai/commit/9942196)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.13 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.39 ### [`v4.0.64`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4064) ##### Patch Changes - [`ccb8952`](https://redirect.github.com/vercel/ai/commit/ccb8952): fix(openai): return an AI SDK error for empty chat completion choices - Updated dependencies \[[`912fb01`](https://redirect.github.com/vercel/ai/commit/912fb01)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.12 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.38 ### [`v4.0.63`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4063) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`85db433`](https://redirect.github.com/vercel/ai/commit/85db433): Include explicit message item types in Azure AI Foundry Responses requests. - [`5fb2a64`](https://redirect.github.com/vercel/ai/commit/5fb2a64): feat(openai): add GPT Image 2.5 Flare and Sunburst model IDs - [`e105b2b`](https://redirect.github.com/vercel/ai/commit/e105b2b): fix(openai): reject denied programmatic tool calls - [`45f2b6a`](https://redirect.github.com/vercel/ai/commit/45f2b6a): fix(openai): report apply patch responses with a tool-calls finish reason ### [`v4.0.62`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4062) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`8487955`](https://redirect.github.com/vercel/ai/commit/8487955): Add `xhigh` and `max` quality support for GPT Image 2.5 Flare and Sunburst in image generation, image editing, and the Responses API image generation tool. ### [`v4.0.61`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4061) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`a4ba394`](https://redirect.github.com/vercel/ai/commit/a4ba394): feat: support per-request models in batch - [`4a09793`](https://redirect.github.com/vercel/ai/commit/4a09793): feat(openai): add async tool calling - [`685ed8c`](https://redirect.github.com/vercel/ai/commit/685ed8c): chore: fix failing CI - Updated dependencies \[[`a4ba394`](https://redirect.github.com/vercel/ai/commit/a4ba394)] - Updated dependencies \[[`45099da`](https://redirect.github.com/vercel/ai/commit/45099da)] - Updated dependencies \[[`9e1d1b2`](https://redirect.github.com/vercel/ai/commit/9e1d1b2)] - Updated dependencies \[[`a495511`](https://redirect.github.com/vercel/ai/commit/a495511)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.11 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.37 ### [`v4.0.60`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4060) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`17e489e`](https://redirect.github.com/vercel/ai/commit/17e489e): feat(openai): add GPT-6 reasoning configuration updates ### [`v4.0.59`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4059) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`4af00d1`](https://redirect.github.com/vercel/ai/commit/4af00d1): feat(openai): add support for the gpt-6-astra - [`abb9ebf`](https://redirect.github.com/vercel/ai/commit/abb9ebf): feat(openai): support `gpt-4o-transcribe-diarize`, including chunking and diarized speaker metadata ### [`v4.0.58`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4058) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`b54e551`](https://redirect.github.com/vercel/ai/commit/b54e551): fix(openai): support documented Responses image generation options - [`17d3436`](https://redirect.github.com/vercel/ai/commit/17d3436): feat(openai): add support for ultrafast service tier ### [`v4.0.57`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4057) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`048ce06`](https://redirect.github.com/vercel/ai/commit/048ce06): feat(batch): surface the uploaded input file on the batch start result (`providerMetadata.<provider>.inputFileId` / `inputFileExpiresAt`) and accept an `inputFileExpiresAfter` provider option on the OpenAI and xAI batch input file upload - [`7243530`](https://redirect.github.com/vercel/ai/commit/7243530): fix(openai): preserve complete Responses API raw usage objects ### [`v4.0.56`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4056) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - Updated dependencies \[[`6bcc0f8`](https://redirect.github.com/vercel/ai/commit/6bcc0f8)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.36 ### [`v4.0.55`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4055) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`3fc40db`](https://redirect.github.com/vercel/ai/commit/3fc40db): feat(openai): implement `getFileMetadata`, `downloadFile` (streaming), and `deleteFile` on the OpenAI files interface, support streaming uploads via `{ type: 'stream' }` data, expose `byteSize`/`createdAt`/`expiresAt` on upload results, and thread `abortSignal`/`headers` through all file operations; blank and dot-segment file ids are rejected/encoded so they cannot retarget request paths - Updated dependencies \[[`5190b67`](https://redirect.github.com/vercel/ai/commit/5190b67)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.10 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.35 ### [`v4.0.54`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4054) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`f6fac50`](https://redirect.github.com/vercel/ai/commit/f6fac50): fix(openai): handle null usage in responses - [`e07b577`](https://redirect.github.com/vercel/ai/commit/e07b577): feat: add tool calling support to batch ### [`v4.0.53`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4053) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`7439d7a`](https://redirect.github.com/vercel/ai/commit/7439d7a): Fix workflow deserialization for OpenAI Responses models. ### [`v4.0.52`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4052) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`1c68540`](https://redirect.github.com/vercel/ai/commit/1c68540): Preserve explicit prompt cache breakpoints on scalar Responses tool results. - Updated dependencies \[[`aa45741`](https://redirect.github.com/vercel/ai/commit/aa45741)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.9 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.34 ### [`v4.0.51`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4051) ##### Patch Changes - Updated dependencies \[[`90192f1`](https://redirect.github.com/vercel/ai/commit/90192f1)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.33 ### [`v4.0.50`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4050) ##### Patch Changes - [`e6a2992`](https://redirect.github.com/vercel/ai/commit/e6a2992): Fix Responses history serialization for regular functions named `tool_search`. ### [`v4.0.49`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4049) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`2523403`](https://redirect.github.com/vercel/ai/commit/2523403): Normalize non-object replayed Chat Completions tool arguments to empty objects. - Updated dependencies \[[`3e125ba`](https://redirect.github.com/vercel/ai/commit/3e125ba)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.32 ### [`v4.0.47`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4047) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`591d25b`](https://redirect.github.com/vercel/ai/commit/591d25b): feat: add batch completion webhooks. `experimental_startTextBatch` accepts a `webhookUrl`, and the gateway provider registers it through the batch `callbackUrl` contract and exports typed async-job metadata. Direct Anthropic and OpenAI batch providers return an unsupported warning when the option is provided. - [`2214258`](https://redirect.github.com/vercel/ai/commit/2214258): Prevent negative text output token counts when providers report reasoning tokens. Perplexity reasoning tokens are now treated as separate from completion tokens. - Updated dependencies \[[`591d25b`](https://redirect.github.com/vercel/ai/commit/591d25b)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.8 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.30 ### [`v4.0.46`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4046) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - Updated dependencies \[[`b74971f`](https://redirect.github.com/vercel/ai/commit/b74971f)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.29 ### [`v4.0.45`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4045) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`6be0f51`](https://redirect.github.com/vercel/ai/commit/6be0f51): fix(openai): expand internal parallel tool call wrappers from the Responses API while preserving stateful continuation and streaming fallbacks ### [`v4.0.44`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4044) ##### Patch Changes - [`e6087c9`](https://redirect.github.com/vercel/ai/commit/e6087c9): fix: handle empty string tool call IDs - Updated dependencies \[[`e6087c9`](https://redirect.github.com/vercel/ai/commit/e6087c9)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.28 ### [`v4.0.43`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4043) ##### Patch Changes - [`a062795`](https://redirect.github.com/vercel/ai/commit/a062795): fix(openai): support built-in and provider-defined tools in the Responses `allowedTools` option `allowedTools` emitted every allow-list entry as `{ type: 'function', name }`, but OpenAI identifies built-in tools by type. Allow-listing a declared provider-defined tool (web search, image generation, MCP, custom, ...) therefore failed with `Tool choice '<name>' not found in 'tools' parameter`. Entries are now derived from the declared tool, including the MCP server label and custom tool name. Tools that OpenAI cannot allow-list (the tool search tool, deferred tools, and namespaced tools) are dropped from the allow-list with a warning, and an error is thrown if that would leave the allow-list empty rather than silently sending an unrestricted request. Ambiguous names are now reported instead of resolved silently. A name that matches both a declared tool and another tool's provider tool name resolves to the declared tool and warns; a provider tool name shared by several tools in the same request (two MCP servers, for example) is dropped with a warning. A name that matches no declared tool keeps its existing behavior and is now warned about. ### [`v4.0.42`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4042) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`b6fff2e`](https://redirect.github.com/vercel/ai/commit/b6fff2e): feat(provider/openai): support explicit Responses compaction triggers ### [`v4.0.41`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4041) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - [`59d6def`](https://redirect.github.com/vercel/ai/commit/59d6def): Reconstruct provider-executed shell calls when continuing OpenAI Responses with storage disabled. ### [`v4.0.40`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4040) ##### Patch Changes - [`e19a4a6`](https://redirect.github.com/vercel/ai/commit/e19a4a6): Avoid duplicate MCP approval request references when continuing stored OpenAI Responses. ### [`v4.0.39`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4039) ##### Patch Changes - Updated dependencies \[[`7fbfc6d`](https://redirect.github.com/vercel/ai/commit/7fbfc6d)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.27 ### [`v4.0.38`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4038) ##### Patch Changes - [`d302134`](https://redirect.github.com/vercel/ai/commit/d302134): Keep client-executed function calls paired with their outputs when chaining OpenAI Responses with a previous response ID. ### [`v4.0.37`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/openai/CHANGELOG.md#4037) [Compare Source](https://redirect.github.com/vercel/ai/compare/@ai-sdk/[email protected]...@ai-sdk/[email protected]) ##### Patch Changes - Updated dependencies \[[`401a4ba`](https://redirect.github.com/vercel/ai/commit/401a4ba)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.26 </details> <details> <summary>vercel/ai (ai)</summary> ### [`v7.0.137`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/ai/CHANGELOG.md#70137) [Compare Source](https://redirect.github.com/vercel/ai/compare/[email protected]@7.0.137) ##### Patch Changes - [`e27cc64`](https://redirect.github.com/vercel/ai/commit/e27cc64): fix(ai): discard unrelated message state when reading resumed streams ### [`v7.0.136`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/ai/CHANGELOG.md#70136) [Compare Source](https://redirect.github.com/vercel/ai/compare/[email protected]@7.0.136) ##### Patch Changes - [`7b5497c`](https://redirect.github.com/vercel/ai/commit/7b5497c): Fix streaming output timeouts: stop `chunkMs` and `firstChunkMs` when the model response ends, so long-running local tools do not trigger model output timeouts. Give each retry fresh output timeout budgets. `stepMs` continues to cover the complete step, including local tools. - Updated dependencies \[[`e280aa1`](https://redirect.github.com/vercel/ai/commit/e280aa1)] - [@​ai-sdk/gateway](https://redirect.github.com/ai-sdk/gateway)@4.0.110 ### [`v7.0.135`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/ai/CHANGELOG.md#70135) [Compare Source](https://redirect.github.com/vercel/ai/compare/[email protected]@7.0.135) ##### Patch Changes - [`5ddbb70`](https://redirect.github.com/vercel/ai/commit/5ddbb70): fix(ai): prevent duplicate continuation requests for repeated tool approval responses - [`7c40a09`](https://redirect.github.com/vercel/ai/commit/7c40a09): feat(ai): log warning when default step limit stops a tool loop - [`4dbc902`](https://redirect.github.com/vercel/ai/commit/4dbc902): feat(ai): add a top level 'dimensions' setting for embeddding functions - [`ff18b73`](https://redirect.github.com/vercel/ai/commit/ff18b73): fix(ai): wait for chat stream cleanup when stopping - [`9be846c`](https://redirect.github.com/vercel/ai/commit/9be846c): feat(ai): add a persistent WebSocket chat transport with correlated streaming, serialized backpressure, cancellation, validated protocol frames, explicit lifecycle, and sequence-based resume support - Updated dependencies \[[`5cf13d2`](https://redirect.github.com/vercel/ai/commit/5cf13d2)] - Updated dependencies \[[`4dbc902`](https://redirect.github.com/vercel/ai/commit/4dbc902)] - Updated dependencies \[[`908fd12`](https://redirect.github.com/vercel/ai/commit/908fd12)] - [@​ai-sdk/gateway](https://redirect.github.com/ai-sdk/gateway)@4.0.109 - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.26 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.58 ### [`v7.0.134`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/ai/CHANGELOG.md#70134) [Compare Source](https://redirect.github.com/vercel/ai/compare/[email protected]@7.0.134) ##### Patch Changes - [`8281ec4`](https://redirect.github.com/vercel/ai/commit/8281ec4): fix(ai): propagate async sendAutomaticallyWhen failures from tool updates - [`46c2b05`](https://redirect.github.com/vercel/ai/commit/46c2b05): fix(ai): preserve provider-executed tool calls when resuming streams ### [`v7.0.133`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/ai/CHANGELOG.md#70133) [Compare Source](https://redirect.github.com/vercel/ai/compare/[email protected]@7.0.133) ##### Patch Changes - [`9184a35`](https://redirect.github.com/vercel/ai/commit/9184a35): Add ordered text, file, and JSON parts to experimental decision state. Support image input in OpenAI Decisions and language model adapters. Gateway retains its existing string and JSON request format and rejects files, as does TypeSafe AI. Arrays passed directly as state now contain decision state parts. Wrap JSON arrays in an object or a json part. Normalize all public state forms into an array of parts before calling decision providers. Providers receive text and JSON objects as text and json parts. Serialize decision file bytes as base64 in OpenTelemetry state attributes. Model-call spans record normalized state parts, while outer spans retain public state inputs. Preserve native JSON state in TypeSafe AI when state contains one JSON part. Label shared state in language-model decision prompts. - Updated dependencies \[[`9184a35`](https://redirect.github.com/vercel/ai/commit/9184a35)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.25 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.57 - [@​ai-sdk/gateway](https://redirect.github.com/ai-sdk/gateway)@4.0.108 ### [`v7.0.132`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/ai/CHANGELOG.md#70132) [Compare Source](https://redirect.github.com/vercel/ai/compare/[email protected]@7.0.132) ##### Patch Changes - [`3ebefff`](https://redirect.github.com/vercel/ai/commit/3ebefff): feat(ai): introduce 'onUIMessageStepEnd' callback to persist messages ### [`v7.0.131`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/ai/CHANGELOG.md#70131) [Compare Source](https://redirect.github.com/vercel/ai/compare/[email protected]@7.0.131) ##### Patch Changes - Updated dependencies \[[`866e884`](https://redirect.github.com/vercel/ai/commit/866e884)] - Updated dependencies \[[`04fdf5e`](https://redirect.github.com/vercel/ai/commit/04fdf5e)] - [@​ai-sdk/gateway](https://redirect.github.com/ai-sdk/gateway)@4.0.107 ### [`v7.0.130`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/ai/CHANGELOG.md#70130) [Compare Source](https://redirect.github.com/vercel/ai/compare/[email protected]@7.0.130) ##### Patch Changes - [`6f6b9c0`](https://redirect.github.com/vercel/ai/commit/6f6b9c0): fix(ai): reject pending result promises and cancel the pending stream when the abort signal fires - [`d3bcad9`](https://redirect.github.com/vercel/ai/commit/d3bcad9): feat(provider): report decision refusals as refusal answers - Updated dependencies \[[`d3bcad9`](https://redirect.github.com/vercel/ai/commit/d3bcad9)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.24 - [@​ai-sdk/gateway](https://redirect.github.com/ai-sdk/gateway)@4.0.106 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.56 ### [`v7.0.129`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/ai/CHANGELOG.md#70129) [Compare Source](https://redirect.github.com/vercel/ai/compare/[email protected]@7.0.129) ##### Patch Changes - [`e37d213`](https://redirect.github.com/vercel/ai/commit/e37d213): feat(ai): emit deprecation warnings with stable codes and deduplicate them - [`73d8343`](https://redirect.github.com/vercel/ai/commit/73d8343): fix(ai): preserve V2 provider-reported total tokens in raw usage - [`268225c`](https://redirect.github.com/vercel/ai/commit/268225c): feat(provider): add a portable `max` reasoning level with native provider mappings, compatibility coercions, and budget-based fallback support. Compatibility notes: - **TypeScript source compatibility:** Adding `max` widens `LanguageModelV4CallOptions['reasoning']` and the derived `ReasoningLevel` type. Third-party providers and consumers with exhaustive switches or `Record<ReasoningLevel, ...>` mappings must handle `max` before upgrading. Ordinary calls using existing reasoning levels remain accepted. - **Amazon Bedrock:** Existing Nova 2 calls using portable `reasoning: 'xhigh'` now send `maxReasoningEffort: 'high'` instead of the unsupported `'max'`, and return a compatibility warning. Anthropic and OpenAI Bedrock effort mappings retain their existing `xhigh` behavior. Partial Bedrock reasoning configurations continue to preserve explicit values and derive missing fields; explicit effort/budget overrides and disabled thinking no longer emit warnings for portable mappings that are not sent. - **Fireworks:** Existing portable `minimal` and `xhigh` calls keep their request mappings to `low` and `high`, respectively, and now return compatibility warnings in generation and streaming results. Portable `max` maps to `high` with the same warning behavior. **Release classification:** This remains a patch changeset under the repository's explicit release policy, which uses patch releases for both fixes and features. The TypeScript source-compatibility caveat for exhaustive consumers is disclosed above. Maintainers can override this classification with a `major` label if they decide to align the provider-spec addition with a future AI SDK major release. - [`d83366e`](https://redirect.github.com/vercel/ai/commit/d83366e): fix(ai): publish the latest error after repeated failed stream resume attempts - [`34905a2`](https://redirect.github.com/vercel/ai/commit/34905a2): fix(ai): preserve provider options from approval-only tool messages - [`16ab882`](https://redirect.github.com/vercel/ai/commit/16ab882): fix(ai): recognize WebKit network errors - [`f810ea3`](https://redirect.github.com/vercel/ai/commit/f810ea3): fix(ai): prevent unavailable tools from exposing persisted output to models - [`19a127b`](https://redirect.github.com/vercel/ai/commit/19a127b): fix(ai): avoid duplicate slashes in chat reconnect URLs - [`0af2f7c`](https://redirect.github.com/vercel/ai/commit/0af2f7c): chore: upgrade to TypeScript 6 - [`4c6979c`](https://redirect.github.com/vercel/ai/commit/4c6979c): fix(harness): preserve adapter warnings in agent stream results and declare their compatibility warning shapes - Updated dependencies \[[`fc1e19e`](https://redirect.github.com/vercel/ai/commit/fc1e19e)] - Updated dependencies \[[`5094ebd`](https://redirect.github.com/vercel/ai/commit/5094ebd)] - Updated dependencies \[[`753f2e1`](https://redirect.github.com/vercel/ai/commit/753f2e1)] - Updated dependencies \[[`268225c`](https://redirect.github.com/vercel/ai/commit/268225c)] - Updated dependencies \[[`0af2f7c`](https://redirect.github.com/vercel/ai/commit/0af2f7c)] - Updated dependencies \[[`ad64697`](https://redirect.github.com/vercel/ai/commit/ad64697)] - Updated dependencies \[[`0c82824`](https://redirect.github.com/vercel/ai/commit/0c82824)] - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.55 - [@​ai-sdk/gateway](https://redirect.github.com/ai-sdk/gateway)@4.0.105 - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.23 ### [`v7.0.128`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/ai/CHANGELOG.md#70128) [Compare Source](https://redirect.github.com/vercel/ai/compare/[email protected]@7.0.128) ##### Patch Changes - [`0fe8c67`](https://redirect.github.com/vercel/ai/commit/0fe8c67): fix: preserve tool approval state when resuming streams - [`2959d35`](https://redirect.github.com/vercel/ai/commit/2959d35): feat: rename `evaluate` to `decide` - [`0ca1ab9`](https://redirect.github.com/vercel/ai/commit/0ca1ab9): Add durable non-streaming `WorkflowAgent.generate()` with a shared tool loop, a 20-step default, typed output, and core generation result semantics. Reuse core result and content construction through internal exports while preserving existing Workflow streaming behavior. Transport-independent approval creation remains separate follow-up work. - [`d6b42fd`](https://redirect.github.com/vercel/ai/commit/d6b42fd): feat(ai): support custom reasoning delimiters in extractReasoningMiddleware - [`ed6e72d`](https://redirect.github.com/vercel/ai/commit/ed6e72d): fix(ai): return successful empty transcripts for silent audio - [`2136151`](https://redirect.github.com/vercel/ai/commit/2136151): Throw `InvalidResponseDataError` instead of a generic `Error` when a generated audio file's format cannot be determined from its media type, so callers can identify the failure with `AISDKError.isInstance`. - Updated dependencies \[[`131532b`](https://redirect.github.com/vercel/ai/commit/131532b)] - Updated dependencies \[[`2959d35`](https://redirect.github.com/vercel/ai/commit/2959d35)] - Updated dependencies \[[`59116e6`](https://redirect.github.com/vercel/ai/commit/59116e6)] - Updated dependencies \[[`6ac923a`](https://redirect.github.com/vercel/ai/commit/6ac923a)] - [@​ai-sdk/gateway](https://redirect.github.com/ai-sdk/gateway)@4.0.104 - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.22 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.54 ### [`v7.0.127`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/ai/CHANGELOG.md#70127) [Compare Source](https://redirect.github.com/vercel/ai/compare/[email protected]@7.0.127) ##### Patch Changes - [`158a718`](https://redirect.github.com/vercel/ai/commit/158a718): feat(ai): select and rank eligible deferred tools via 'search()' callback in tool search - [`bb8d33e`](https://redirect.github.com/vercel/ai/commit/bb8d33e): fix(ai): cancel merged UI message streams when the consumer disconnects - [`284dc11`](https://redirect.github.com/vercel/ai/commit/284dc11): fix(ai): accept unchanged tool approval inputs created in another JavaScript realm - [`ba8afe9`](https://redirect.github.com/vercel/ai/commit/ba8afe9): feat(ai): add a configurable maxResults for number of tools returned in tool search - Updated dependencies \[[`d1bb9e8`](https://redirect.github.com/vercel/ai/commit/d1bb9e8)] - [@​ai-sdk/gateway](https://redirect.github.com/ai-sdk/gateway)@4.0.103 ### [`v7.0.126`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/ai/CHANGELOG.md#70126) [Compare Source](https://redirect.github.com/vercel/ai/compare/[email protected]@7.0.126) ##### Patch Changes - [`4f3d236`](https://redirect.github.com/vercel/ai/commit/4f3d236): fix: clear tool approvals when `addToolOutput` runs ### [`v7.0.124`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/ai/CHANGELOG.md#70124) [Compare Source](https://redirect.github.com/vercel/ai/compare/[email protected]@7.0.124) ##### Patch Changes - [`8c65988`](https://redirect.github.com/vercel/ai/commit/8c65988): feat(ai): add telemetry to speech generation and transcription, including provider usage propagation and experimental streaming lifecycle callbacks - Updated dependencies \[[`8c65988`](https://redirect.github.com/vercel/ai/commit/8c65988)] - Updated dependencies \[[`e25994e`](https://redirect.github.com/vercel/ai/commit/e25994e)] - Updated dependencies \[[`527a163`](https://redirect.github.com/vercel/ai/commit/527a163)] - Updated dependencies \[[`e74afc5`](https://redirect.github.com/vercel/ai/commit/e74afc5)] - [@​ai-sdk/provider](https://redirect.github.com/ai-sdk/provider)@4.0.21 - [@​ai-sdk/provider-utils](https://redirect.github.com/ai-sdk/provider-utils)@5.0.53 - [@​ai-sdk/gateway](https://redirect.github.com/ai-sdk/gateway)@4.0.102 ### [`v7.0.123`](https://redirect.github.com/vercel/ai/blob/HEAD/packages/ai/CHANGELOG.md#70123) [Compare Source](https://redirect.github.com/vercel/ai/compare/[email protected]@7.0.123) ##### Patch Changes - [`05cdac6`](https://redirect.github.com/vercel/ai/commit/05cdac6): fix(ai): keep idle UI message streams open with optional SSE heartbeats - [`4514fc1`](https://redirect.github.com/vercel/ai/commit/4514fc1): fix(ai): prune all tool content when retaining zero trailing messages - [`ede5b89`](https://redirect.github.com/vercel/ai/commit/ede5b89): chore: migrate package builds from tsup to tsdown - [`2a625cf`](https://redirect.github.com/vercel/ai/commit/2a625cf): fix(ai): preserve partial reasoning tags when streamed text parts end - [`50a26d5`](https://redirect.github.com/vercel/ai/commit/50a26d5): fix(ai): ignore pending tool approvals superseded by user messages - Updated dependencies \[[`040033b`](https://redirect.github.com/vercel/ai/commit/040033b)] - Updated > ❗ **Important** > > ✂ PR body was truncated to here. </details> --- ### Configuration 📅 **Schedule**: (in timezone Etc/UTC) - Branch creation - "before 8am on monday" - Automerge - At any time (no schedule defined) 🚦 **Automerge**: Disabled by config. Please merge this manually once you are satisfied. ♻ **Rebasing**: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox. 👻 **Immortal**: This PR will be recreated if closed unmerged. Get [config help](https://redirect.github.com/renovatebot/renovate/discussions) if that's undesired. --- - [ ] <!-- rebase-check -->If you want to rebase/retry this PR, check this box --- This PR was generated by [Mend Renovate](https://mend.io/renovate/). View the [repository job log](https://developer.mend.io/github/apache/texera). <!--renovate-debug:eyJjcmVhdGVkSW5WZXIiOiI0NC4xMzQuMSIsInVwZGF0ZWRJblZlciI6IjQ0LjEzNC4xIiwidGFyZ2V0QnJhbmNoIjoibWFpbiIsImxhYmVscyI6WyJkZXBlbmRlbmNpZXMiXX0=--> -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
