NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
npm
npm, #4482 most downloaded on npm
Latest stable
4.0.102
Last release 12 days ago
10 Nov 2026
Ships on a steady schedule
a new release about every 9 days
Most releases are documented
notes for 36 of the last 60 stable releases
Nothing withdrawn
no release was ever pulled
3 years old
688 releases · first in 2024
Drag across the ruler to compare, or tap a release to open it.
### Patch Changes - Updated dependencies [4311fa1] - Updated dependencies [4311fa1] - @ai-sdk/provider-utils@5.0.79
### Patch Changes - Updated dependencies [88b9575] - @ai-sdk/provider-utils@5.0.78
a161c0a: Keep computer toolset actions tied to the returned member name and reject disabled or unavailable members before emitting tool calls. Apply m
resolveProviderOptions to parse provider options that a model accepts under several providerOptions keys, either as aliases (first) or merged in order (merge), and use it for the existing multi-key lookups in the OpenAI, OpenAI-compatible, Anthropic, Google, and Amazon Bedrock models. Resolution behavior for these models is unchanged.Your coding agent can read these notes before it upgrades. Set up the MCP server →
067dd77: Refactor Amazon Bedrock model ID logic into a model catalog that declares capabilities and provider options, and generate provider capability
toolChanges (inline-tools-2026-09-15)fd7e8a7: Preserve the order of text and complete function calls in streamed Google responses, including standalone thought signatures on empty text or
fd7e8a7: Preserve the order of text and complete function calls in streamed Google responses, including standalone thought signatures on empty text or reasoning parts.
Keep provider metadata on empty text parts when converting results to response messages so it can be passed back in subsequent requests.
Omit foreign-provider empty signature text when replaying history through Anthropic or Bedrock, preserving Bedrock’s existing native reasoning-history handling.
8291ccf: Add MiniMax serviceTier support for standard and priority admission. Omit tiers belonging to another known provider with an unsupported-setting warning instead of forwarding incompatible values. Expose the applied serviceTier, when returned, in Anthropic-compatible generation, stream and batch metadata without changing existing metadata construction types.
9754014: Add typed streaming heartbeat parts. Anthropic emits them only with includePings: true (disabled by default), without requiring raw chunks. Core forwards pings through stream retries without storing them as generated content or sending them to UI message streams. Consumers with exhaustive stream-part switches must handle the new ping union variant when upgrading, even when the opt-in is disabled. Typed pings do not flush partial words during stream smoothing.
4b8f00e: Add a supportsNativeStructuredOutput opt-in to the Vertex Anthropic provider so auto can use native structured output for Claude 4.5 and newer when enabled by Google Cloud policy. Keep tool emulation as the default and avoid adding the structured-output beta header for providers without strict tool support.
8b11ff6: Support deleting files, retrieving metadata, and streaming downloads of generated files through the Anthropic Files API. Metadata reads and uploads expose consistent media type, size, creation time, expiration, and provider metadata when available. Downloads preserve Anthropic’s restriction to files generated by tools or skills.
Updated dependencies [36802d4]
Updated dependencies [e0143b0]
Updated dependencies [934100c]
Updated dependencies [892af4d]
Updated dependencies [89aff56]
Updated dependencies [41465fd]
Updated dependencies [9754014]
Updated dependencies [4bb73b0]
Updated dependencies [2e44144]
### Patch Changes - Updated dependencies [fa15312] - @ai-sdk/provider-utils@5.0.74
2cd0a97: Enable native structured output and strict tools for supported Claude 5 models on Bedrock. In automatic mode, Bedrock and Anthropic now fall
outputFormat selection remains unchanged.f4c5f1e: fix(anthropic): use native structured outputs for Claude models that reject forced tool use (Sonnet 5.5, Opus 5.5, Fable 5.1) even when the p
definitions and $defs next to a top-level $ref when sanitizing structured output schemas, so the reference still resolves.auto), return the final answer text instead of dropping it. Text before a tool call, result, or message continuation is still dropped.{}, and schemas with more than 10 optional properties can be rejected as too complex.dcd2f8d: Introduce a model entry contract with named capability sections and request configuration. Consolidate the OpenAI and Anthropic model catalog
61a315e: fix: map the top-level reasoning option consistently across providers
61a315e: fix: map the top-level reasoning option consistently across providers
Providers now resolve reasoning the same way:
providerOptions is sent unchanged. OpenAI no longer drops providerOptions.openai.reasoningEffort values that are not in its list for the model, and Fireworks no longer rewrites providerOptions.fireworks.reasoningEffort. Fireworks also sends top-level reasoning values directly instead of rounding them to 'high' at most, except for model families known to reject 'none' (GPT-OSS, GLM-5.3, Muse Spark) or 'minimal' (Nemotron, Inkling).reasoning maps to the closest level the model supports, rounding down, and rounds up only when there is no lower level. For example, 'xhigh' now maps to 'high' instead of 'max' on Anthropic and Bedrock models without xhigh effort, and 'medium' maps to 'low' instead of 'high' on DeepSeek, Kimi K3, and Mistral GLM-5.3 models.'none' on a model that cannot disable reasoning maps to the model's lowest level with a compatibility warning, instead of being dropped or sending a request the API rejects. This affects GPT-6 models without 'none', xAI grok-4.5 and later, Groq GPT-OSS, Kimi K3, Perplexity, QuiverAI, Bedrock OpenAI gpt-oss models, and Gemini 2.5 Pro (minimum thinking budget). Gemini 3 and Anthropic and Bedrock between_tools mappings now emit the compatibility warning too.mapReasoningToProviderEffort now accepts any reasoning value, returns an optional providerOptionsEffort unchanged, and rounds levels that are missing from the effort map to the closest supported level. mapReasoningToProviderBudget maps 'none' to the minimum reasoning budget, always returns a budget, and uses the default percentage for levels missing from budgetPercentages. OpenAICompatibleChatConfig adds an optional mapReasoning function for mapping the top-level reasoning option.
Updated dependencies [61a315e]
50da4c5: Fix workflow serialization of Anthropic models using API keys or synchronous auth tokens while preserving support for asynchronous token prov
63bf860: Support synchronous and asynchronous authToken providers so applications can use refreshable credentials, including Anthropic workload identi
authToken providers so applications can use refreshable credentials, including Anthropic workload identity federation helpers for inference. Resolve tokens before each request while preserving existing authentication headers and beta-header behavior.### Patch Changes - Updated dependencies [e64ed82] - @ai-sdk/provider-utils@5.0.69
e729d58: chore(docs): update README.md for consistent format and to avoid duplication
README.md for consistent format and to avoid duplicationf5fc6ba: Derive known model ID suggestions from each provider's model catalog declarations so catalog updates automatically update autocomplete, while
### Patch Changes - Updated dependencies [6d50947] - @ai-sdk/provider-utils@5.0.81
43ad966: Report interrupted when a stream ends cleanly without its provider completion signal, preserving partial output without emitting an error solely for the interruption.
This applies to OpenAI (Chat, Responses, and Completions), Anthropic, Google (Generate Content and Interactions), Alibaba, Amazon Bedrock Converse, Cohere, DeepSeek, Groq, Hugging Face Responses, Mistral Chat, Moonshot AI, Open Responses, Perplexity, and xAI Responses. It also covers cutoffs in later Anthropic messages and Google automatic continuations.
Compatibility notice: These streams previously fell back to other or omitted the finish part. Applications must now handle interrupted in finish-reason switches, mappings, and custom validation. Check finishReason === 'interrupted' to detect incomplete responses; interruption alone does not trigger error handling.
Explicit unknown completion reasons still map to other. Actual provider, parsing, and transport errors retain their error paths; affected error paths that previously left the finish reason as other now report error. OpenAI-compatible remains unchanged.
aac6026: fix(anthropic): preserve empty-stream rejection for interrupted responses
Updated dependencies [10026d1]
### Patch Changes - Updated dependencies [2ebc78b] - Updated dependencies [cb4fa52] - @ai-sdk/provider-utils@5.0.80 - @ai-sdk/provider@4.0.38
f5fc6ba: Derive known model ID suggestions from each provider's model catalog declarations so catalog updates automatically update autocomplete, while
### Patch Changes - Updated dependencies [2ebc78b] - Updated dependencies [cb4fa52] - @ai-sdk/provider-utils@5.0.80 - @ai-sdk/provider@4.0.38
### Patch Changes - Updated dependencies [2ebc78b] - Updated dependencies [cb4fa52] - @ai-sdk/provider-utils@5.0.80 - @ai-sdk/provider@4.0.38
e254e96: fix(anthropic): support provider-referenced files in tool results
40d0aec: feat(vertex-anthropic): forward 'providerOptions.vertex.labels' as a header
1ccc7fb: fix(anthropic): preserve tools when disabling tool calls
### Patch Changes - Updated dependencies [4dbc902] - @ai-sdk/provider@4.0.26 - @ai-sdk/provider-utils@5.0.58
c01cbb8: fix(anthropic): avoid unsupported strict mode warnings when tools set strict: false
strict: false### Patch Changes - Updated dependencies [9184a35] - @ai-sdk/provider@4.0.25 - @ai-sdk/provider-utils@5.0.57
866e884: feat(anthropic): add Claude Haiku 5.5 support
866e884: feat(anthropic): add Claude Haiku 5.5 support
claude-haiku-5-5 model ID to @ai-sdk/anthropic and @ai-sdk/google-vertex, anthropic.claude-haiku-5-5 and us.anthropic.claude-haiku-5-5 to @ai-sdk/amazon-bedrock, and anthropic/claude-haiku-5.5 to @ai-sdk/gatewayclaude-haiku-5-5 as a known model with a 128k max output token limit, adaptive thinking, and all five effort levels; thinking can be disabled up to high effort, and xhigh and max are lowered to high with a warning when thinking is disabledclaude-haiku-5-5 does not support thinking token budgets: thinking: { type: 'enabled', budgetTokens } is converted to adaptive thinking with a warningclaude-haiku-5-5### Patch Changes - Updated dependencies [d3bcad9] - @ai-sdk/provider@4.0.24 - @ai-sdk/provider-utils@5.0.56
fc1e19e: fix: limit jsonl rows to 64mb during batch result downloads
fc1e19e: fix: limit jsonl rows to 64mb during batch result downloads
268225c: feat(provider): add a portable max reasoning level with native provider mappings, compatibility coercions, and budget-based fallback support.
Compatibility notes:
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.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.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: chore: upgrade to TypeScript 6
Updated dependencies [fc1e19e]
Updated dependencies [5094ebd]
Updated dependencies [268225c]
Updated dependencies [0af2f7c]
2959d35: feat: rename evaluate to decide
evaluate to decidec35458e: fix(anthropic): preserve invalid toolset calls
### Patch Changes - Updated dependencies [8c65988] - Updated dependencies [527a163] - @ai-sdk/provider@4.0.21 - @ai-sdk/provider-utils@5.0.53
ede5b89: chore: migrate package builds from tsup to tsdown
c2511c1: fix: use standards-compliant User-Agent header
e361d39: feat(anthropic): add Claude Sonnet 5.5 support
e361d39: feat(anthropic): add Claude Sonnet 5.5 support
claude-sonnet-5-5 model ID to @ai-sdk/anthropic and @ai-sdk/google-vertex, anthropic.claude-sonnet-5-5 and us.anthropic.claude-sonnet-5-5 to @ai-sdk/amazon-bedrock, and anthropic/claude-sonnet-5.5 to @ai-sdk/gatewaybetween_tools thinking type (thinking: { type: 'between_tools' }), the lowest thinking setting on claude-sonnet-5-5; xhigh and max effort are lowered to high with a warning because the API rejects them with between_toolsclaude-sonnet-5-5 rejects disabled thinking: thinking: { type: 'disabled' } is replaced with between_tools thinking (with a warning), reasoning: 'none' maps to between_tools thinking, and budget-based thinking is converted to adaptive thinkingclaude-sonnet-5-5 rejects forced tool use: required and named tool choices fall back to auto, and structuredOutputMode: 'jsonTool' falls back to native structured outputs, each with a warning8373a22: Normalize dangling programmatic tool caller references in conversation history after pruning. Before a subsequent user message, omit caller metadata whose source code execution call is missing and emit a warning, preserving the retained tool calls and results. Keep caller metadata unchanged for active tool continuations.
### Patch Changes - Updated dependencies [e3605f6] - Updated dependencies [525efc5] - @ai-sdk/provider-utils@5.0.50 - @ai-sdk/provider@4.0.19
### Patch Changes - Updated dependencies [af9597b] - Updated dependencies [bc49f78] - @ai-sdk/provider-utils@5.0.49
154221f: feat(anthropic): support on-demand compaction and preserve signed compaction blocks
771e74b: chore: enable dead code lint rules
3733d6e: fix(anthropic): omit empty compaction blocks from replay
49295bb: feat(anthropic): add Claude Opus 5.5 support
49295bb: feat(anthropic): add Claude Opus 5.5 support
claude-opus-5-5 model ID to @ai-sdk/anthropic and anthropic/claude-opus-5.5 to @ai-sdk/gatewayclaude-opus-5-5, claude-fable-5, claude-fable-5-1) no longer receive thinking: { type: 'disabled' } or budget-based thinking; the provider drops the unsupported setting, maps reasoning: 'none' to effort: 'low', and emits a warningclaude-opus-5-5, claude-fable-5-1) fall back to auto tool choice for required and named tool choices, and to native structured outputs when structuredOutputMode: 'jsonTool' is requested, each with a warningcomputerToolset_20260801 computer use tool (computer_toolset_20260801), which is required for computer use on claude-opus-5-5mid-conversation-output-config-2026-07-01 beta header for per-message effortf7b7b2a: feat(provider/anthropic): add safeguards provider option and safeguardResults provider metadata (dangerous tool use classifier)
safeguards provider option and safeguardResults provider metadata (dangerous tool use classifier)### Patch Changes - Updated dependencies [2973485] - Updated dependencies [a4db5ea] - Updated dependencies [2937ea2] - @ai-sdk/provider-utils@5.0.45
### Patch Changes - Updated dependencies [0455398] - @ai-sdk/provider-utils@5.0.44
d4d96bf: Add anthropic.evaluationModel() for experimental Choice, Score, and Boolean evaluations through Messages structured output, with exact labels
anthropic.evaluationModel() for experimental Choice, Score, and Boolean evaluations through Messages structured output, with exact labels, validated score bounds, and prompted Boolean P(true) estimates validated to be in [0, 1]. Boolean estimates are not guaranteed to be calibrated; application code chooses thresholds.132bdae: feat(anthropic): add 20260318 web search and web fetch tools with response inclusion controls
### Patch Changes - Updated dependencies [5c0054d] - Updated dependencies [39535af] - @ai-sdk/provider@4.0.15 - @ai-sdk/provider-utils@5.0.41
5ec21a6: fix: reject unsupported batch request types
prefix_mismatch_behavior: 'error'5a7e647: feat(anthropic): add batch cancellation and listing
### Patch Changes - Updated dependencies [912fb01] - @ai-sdk/provider@4.0.12 - @ai-sdk/provider-utils@5.0.38
a4ba394: feat: support per-request models in batch
65397d7: fix(anthropic): recognize dated Google Vertex Claude 4 model IDs when selecting model capabilities
4d25a08: feat(anthropic): add fable 5.1 support
5190b67: feat(provider): extend the FilesV4 interface with optional getFileMetadata, downloadFile (streaming), and deleteFile operations, plus abortSi
getFileMetadata, downloadFile (streaming), and deleteFile operations, plus abortSignal/headers call options and a { type: 'stream' } upload data variant; upload results now expose byteSize, createdAt, and expiresAt (also surfaced by the core uploadFile() helper, which now forwards abortSignal/headers); add postMultipartStreamToApi (streaming multipart uploads with deterministic part ordering and failure-path stream teardown), deleteFromApi, and createBinaryStreamResponseHandler to provider-utilsaa45741: fix(provider/anthropic): preserve native message batch request counts in provider metadata and support the full language-model option surface
### Patch Changes - Updated dependencies [90192f1] - @ai-sdk/provider-utils@5.0.33
### Patch Changes - Updated dependencies [3e125ba] - @ai-sdk/provider-utils@5.0.32
591d25b: feat: add batch completion webhooks. experimental_startTextBatch accepts a webhookUrl, and the gateway provider registers it through the batc
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.### Patch Changes - Updated dependencies [b74971f] - @ai-sdk/provider-utils@5.0.29
### Patch Changes - Updated dependencies [e6087c9] - @ai-sdk/provider-utils@5.0.28
One column per month.