NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
npm
npm, #2521 most downloaded on npm
Latest stable
4.0.146
Last release 8 days ago
13 Nov 2026
Ships on a steady schedule
a new release about every 8 days
Some releases are documented
notes for 29 of the last 60 stable releases
Nothing withdrawn
no release was ever pulled
1 years old
846 releases · first in 2025
Drag across the ruler to compare, or tap a release to open it.
9baa2d0: chore(provider/gateway): update gateway model settings files
### Patch Changes - Updated dependencies [6d50947] - @ai-sdk/provider-utils@5.0.81
640569d: chore(provider/gateway): update gateway model settings files
gemini-4-hydrogen model ID and support Hydrogen's xhigh thinking level in Google Generate Content and Interactions provider options. Map portable reasoning: 'xhigh' to xhigh only for Hydrogen models.Your coding agent can read these notes before it upgrades. Set up the MCP server →
eaeb37e: chore(provider/gateway): update gateway model settings files
838dd53: fix(provider/gateway): restore export_pricing.free_rows on takoSearch results as a deprecated optional field, so code that reads it compiles…
gemini-4-carbon model with the same capabilities as gemini-4-argon.export_pricing.free_rows on takoSearch results as a deprecated optional field, so code that reads it compiles again. Tako always sends 0 and will stop sending it.5497ff7: chore(provider/gateway): update gateway model settings files
44db225: feat(provider/gateway): type the headquarters provider option
### Patch Changes - Updated dependencies [6400cf8] - Updated dependencies [f608c05] - Updated dependencies [cab9c1f] - Updated dependencies [c70adf6]
bace553: chore(provider/gateway): update gateway model settings files
apiKey JSDoc and README722d4a9: chore(provider/gateway): update gateway model settings files
722d4a9: chore(provider/gateway): update gateway model settings files
acea2b5: chore(provider/gateway): update gateway model settings files
89aff56: Add opt-in streaming decoding of large generated images to reduce peak memory.
Set providerOptions.gateway.experimental_streamingImageDecoding to true to decode canonical base64 images of at least 64 KiB while reading the response. Extracted images are returned as Uint8Array; smaller images stay base64 strings, and mixed results are normalized to bytes. The client option is not forwarded to the service.
Default behavior preserves the API's base64 representation. Opting in changes the model result representation and replaces large image data with placeholders in validation errors. Accessing GeneratedFile.base64 can require re-encoding the extracted bytes.
41465fd: Preserve optional normalized word timings returned by transcription responses and warn when shared timestampGranularities requests are not forwarded by the gateway adapter.
2e44144: Accept explicit uploaded-video outcomes in video generation and status responses.
ef2dfb5: fix: pass through seed: 0 instead of dropping it in image and video models. KlingAI now reports seed: 0 as an unsupported setting like any other seed.
892af4d: feat (provider/gateway): implement server-side realtime connection descriptors
GatewayRealtimeModel now implements getServerWebSocketConfig(), returning the AI Gateway realtime WebSocket URL plus Authorization header auth and optional team scoping for trusted server-side clients. Unsupported endpoint intents are rejected on the normalized realtime route.
Server WebSockets reuse HTTP authentication, protocol, team and custom headers, including OIDC auth-method metadata.
32acf43: Add marketCost, surchargeCost, and gatewayCost fields to getGenerationInfo() results, surfacing the corresponding fields from the AI Gateway generation lookup API.
The cost breakdown fields are optional and omitted when unavailable, preserving compatibility with older responses and distinguishing missing data from a zero cost.
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
### Patch Changes - Updated dependencies [51ca5b7] - Updated dependencies [dc6f1c3] - Updated dependencies [4ae6986] - @ai-sdk/provider-utils@5.0.73
2099d45: feat(provider/gateway): add DeepSeek Search and Fetch tools
2099d45: feat(provider/gateway): add DeepSeek Search and Fetch tools
Share lazy tool-error validation across DeepSeek, Parallel, and Tako while preserving additional error fields.
Tako tool errors also accept gateway_rate_limit and unknown; exhaustive error switches may need to handle these cases. Parallel and Tako errors gain optional retryAfter and statusText fields.
d05ff62: chore(provider/gateway): update gateway model settings files
cbc9ac0: chore(provider/gateway): update gateway model settings files
7716e18: Support promises and synchronous or asynchronous callbacks in provider API key and token settings. Resolve credentials per authenticated requ
7716e18: Support promises and synchronous or asynchronous callbacks in provider API key and token settings. Resolve credentials per authenticated request while preserving existing authentication behavior and synchronous credential serialization.
The experimental OpenAI Live getServerWebSocketConfig() method can now resolve asynchronously. Await its result before using the connection settings.
f850782: chore(provider/gateway): update gateway model settings files
Updated dependencies [7716e18]
Updated dependencies [5c247e9]
Updated dependencies [dcd2f8d]
9ca8bcf: fix(gateway): Preserve AbortError and TimeoutError cancellation reasons instead of wrapping them in retryable GatewayResponseError errors wit
9f19c79: chore(provider/gateway): update gateway model settings files
01d7a03: feat(openai): add GPT-6.1 Terra model support
### Patch Changes - Updated dependencies [e64ed82] - @ai-sdk/provider-utils@5.0.69
### Patch Changes - Updated dependencies [f5fc6ba] - @ai-sdk/deepseek@3.0.87 - @ai-sdk/openai@4.0.118
README.md for consistent format and to avoid duplication### Patch Changes - Updated dependencies [6d50947] - @ai-sdk/provider-utils@5.0.81 - @ai-sdk/openai@4.0.117 - @ai-sdk/deepseek@3.0.86
2ebc78b: Forward the provider's configured fetch transport when downloading file inputs. Add a fetch option to createDownload so core downloads can al
2ebc78b: Forward the provider's configured fetch transport when downloading file inputs. Add a fetch option to createDownload so core downloads can also use guarded-fetch for opt-in DNS validation on Node.js. The function returned by createDownload can now also be passed as experimental_download to generateText and streamText, and experimental_download now also handles generated file URLs returned by the model, so one hook covers all core downloads. Document that the default URL checks do not resolve DNS and recommend the guarded transport or network egress controls for untrusted URLs.
Keep Azure image and Vertex tool-result downloads separate from provider authentication wrappers so file hosts never receive Azure bearer tokens or Vertex Express Mode API keys.
Pass the call's abort signal to custom download requests, including prompt assets and generated files, so experimental_download: createDownload({ fetch }) preserves cancellation and timeouts.
Updated dependencies [2ebc78b]
Updated dependencies [2ebc78b]
Updated dependencies [6eff958]
Updated dependencies [a279e86]
Updated dependencies [cb4fa52]
8202f5b: feat(gateway): implement experimental streaming speech in the gateway provider
8202f5b: feat(gateway): implement experimental streaming speech in the gateway provider
GatewaySpeechModel implements Experimental_StreamingSpeechModelV4 and streams speech over the AI Gateway speech WebSocket (/speech-model, ai-gateway-speech.v1 subprotocol). @ai-sdk/provider-utils adds the experimental speech-stream envelope (experimental_parseSpeechStreamClientFrame, experimental_serializeSpeechStreamPart, experimental_parseSpeechStreamPart) shared by clients and servers, plus the EXPERIMENTAL_SPEECH_STREAM_* frame type constants and Experimental_SpeechStream* frame types. Text-delta frames carry each text part's voice and providerOptions; the start frame carries outputFormat as an audio format object. Adds gateway.experimental_speech, callable like speech(modelId), with a server-side getToken() that mints a speech-bound short-lived client secret for browser use.
Updated dependencies [8202f5b]
Updated dependencies [85d84a1]
Tool calls that use one of these modes no longer fail client-side input validation. one-shot and agentic are deprecated but still accepted.
101ce42: chore(provider/gateway): update gateway model settings files
e104368: Add Voxtral 3 offline transcription and speech model IDs. Support transcription context phrases and detected language, update timestamp validation for Voxtral 3, and expose speech expressiveness and pronunciation hints.
eb1f056: fix(provider/gateway): update parallelSearch for Parallel's GA Search API
turbo, fast, basic, and advanced modes. Tool calls that use one of these modes no longer fail client-side input validation. one-shot and agentic are deprecated but still accepted.location and top-level maxCharsTotal options, plus fetchPolicy.timeoutSeconds and fetchPolicy.disableCacheFallback. excerpts.maxCharsTotal is deprecated in favor of maxCharsTotal.search_id, results with url, nullable title and publish_date, and excerpts: string[], plus optional warnings, usage, and session_id. The searchId, excerpt, publishDate, and relevanceScore fields were never populated and are removed; update code that reads them. validateUIMessages now accepts Parallel Search results.5893e99: chore(provider/gateway): update gateway model settings files
Updated dependencies [10026d1]
f5fc6ba: Derive known model ID suggestions from each provider's model catalog declarations so catalog updates automatically update autocomplete, while
reasoning in generation and streaming calls. Map reasoning to model-supported thinking levels, including Hydrogen's xhigh, preserve explicit thinkingLevel overrides and API defaults, and warn when reasoning is unsupported.6d50947: add experimental extraction models backed by structured language model output
2ebc78b: Forward the provider's configured fetch transport when downloading file inputs. Add a fetch option to createDownload so core downloads can al
2ebc78b: Forward the provider's configured fetch transport when downloading file inputs. Add a fetch option to createDownload so core downloads can also use guarded-fetch for opt-in DNS validation on Node.js. The function returned by createDownload can now also be passed as experimental_download to generateText and streamText, and experimental_download now also handles generated file URLs returned by the model, so one hook covers all core downloads. Document that the default URL checks do not resolve DNS and recommend the guarded transport or network egress controls for untrusted URLs.
Keep Azure image and Vertex tool-result downloads separate from provider authentication wrappers so file hosts never receive Azure bearer tokens or Vertex Express Mode API keys.
Pass the call's abort signal to custom download requests, including prompt assets and generated files, so experimental_download: createDownload({ fetch }) preserves cancellation and timeouts.
4e43721: Add the gemini-4-hydrogen model ID and support Hydrogen's xhigh thinking level in Google Generate Content and Interactions provider options. Map portable reasoning: 'xhigh' to xhigh only for Hydrogen models.
Updated dependencies [2ebc78b]
Updated dependencies [cb4fa52]
maxOutputTokens and reasoning to experimental decision calls and forward them to language-model-backed decisions. Language-model adapters keep reasoning disabled when omitted. Native decision providers warn when either setting is unsupported.f5fc6ba: Derive known model ID suggestions from each provider's model catalog declarations so catalog updates automatically update autocomplete, while
6d50947: add experimental extraction models backed by structured language model output
2ebc78b: Forward the provider's configured fetch transport when downloading file inputs. Add a fetch option to createDownload so core downloads can al
2ebc78b: Forward the provider's configured fetch transport when downloading file inputs. Add a fetch option to createDownload so core downloads can also use guarded-fetch for opt-in DNS validation on Node.js. The function returned by createDownload can now also be passed as experimental_download to generateText and streamText, and experimental_download now also handles generated file URLs returned by the model, so one hook covers all core downloads. Document that the default URL checks do not resolve DNS and recommend the guarded transport or network egress controls for untrusted URLs.
Keep Azure image and Vertex tool-result downloads separate from provider authentication wrappers so file hosts never receive Azure bearer tokens or Vertex Express Mode API keys.
Pass the call's abort signal to custom download requests, including prompt assets and generated files, so experimental_download: createDownload({ fetch }) preserves cancellation and timeouts.
6eff958: feat(openai): add positioned additionalTools and accept custom and namespace tools
a279e86: fix(openai): keep tool-result additionalTools for parallel results and openai options
Updated dependencies [2ebc78b]
Updated dependencies [cb4fa52]
5a645e9: Support positioned Responses API format updates so changing structured output schemas can preserve earlier reasoning and prompt cache reuse.
b658e85: chore(provider/gateway): update gateway model settings files
42a2895: chore(provider/gateway): update gateway model settings files
685667c: chore(provider/gateway): update gateway model settings files
gemini-4-argon model id and long decode support for generateContent: requests that reach the per-request cap (CONTINUATION) are continued automatically with usage summed across requests and the first request's prompt token count exposed as providerMetadata.google.initialPromptTokenCount, and the continuationToken provider option / provider metadata resume responses that used up their maxOutputTokens budget. generateText with the Interactions API now polls background: true model calls until they finish.### Patch Changes - Updated dependencies [9667bf2] - Updated dependencies [e57803c] - Updated dependencies [81663cc] - @ai-sdk/provider-utils@5.0.59 -
e280aa1: feat(gateway): send image files in decision state
5cf13d2: feat: map the top level embedding dimensions to provider settings
9184a35: Add ordered text, file, and JSON parts to experimental decision state. Support image input in OpenAI Decisions and language model adapters. G
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]
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-5-504fdf5e: chore(provider/gateway): update gateway model settings files
d3bcad9: feat(provider): report decision refusals as refusal answers
f5fc6ba: Derive known model ID suggestions from each provider's model catalog declarations so catalog updates automatically update autocomplete, while
6d50947: add experimental extraction models backed by structured language model output
evaluate to decide### Patch Changes - Updated dependencies [2ebc78b] - Updated dependencies [cb4fa52] - @ai-sdk/provider-utils@5.0.80 - @ai-sdk/provider@4.0.38
8c65988: feat(ai): add telemetry to speech generation and transcription, including provider usage propagation and experimental streaming lifecycle cal
8c65988: feat(ai): add telemetry to speech generation and transcription, including provider usage propagation and experimental streaming lifecycle callbacks
e25994e: fix(gateway): scope getCredits() to teamIdOrSlug
The credits endpoint reads the team from the teamId / slug query parameter rather than the x-vercel-ai-gateway-team header. getCredits() now forwards the configured team as a query parameter, so credentials that can access multiple teams (such as Vercel access tokens) are no longer rejected with an authentication error.
e74afc5: chore(provider/gateway): update gateway model settings files
Updated dependencies [8c65988]
Updated dependencies [527a163]
040033b: feat(openai): add GPT-6.1 Sol model support
870f509: chore(provider/gateway): update gateway model settings files
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 warning### Patch Changes - Updated dependencies [e3605f6] - Updated dependencies [525efc5] - @ai-sdk/provider-utils@5.0.50 - @ai-sdk/provider@4.0.19
e6a7996: fix(provider/gateway): accept any non-empty conditional evaluation fallback model
b67b1b7: feat(provider/gateway): type conditional evaluation fallbacks
### Patch Changes - Updated dependencies [af9597b] - Updated dependencies [bc49f78] - @ai-sdk/provider-utils@5.0.49
124b6ef: chore(provider/gateway): update gateway model settings files
2693319: Add Gemini 3.8 TTS support with structured speech metadata and per-turn speaker and style controls for prebuilt voices. Preserve native WAV r
2693319: Add Gemini 3.8 TTS support with structured speech metadata and per-turn speaker and style controls for prebuilt voices. Preserve native WAV responses without adding a second header, support explicit raw PCM, mu-law, and A-law output, and identify headerless audio formats correctly. Add the Gemini 3.8 speech model IDs to Google and Gateway types.
Share transcript and custom-voice inspection through the Google provider internal export, and reject empty speech transcripts before sending a request. Default newer and custom model IDs to structured speech while preserving the legacy format for Gemini 2.5 and 3.1.
b73f2f9: feat(provider/gateway): add quantization conditions to the has provider option
Updated dependencies [fe07867]
Updated dependencies [a4b0940]
Updated dependencies [771e74b]
ed5a1d7: chore(provider/gateway): update gateway model settings files
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 effort20dd00a: feat(provider/gateway): support reasoning and tool-use in has model filtering
reasoning and tool-use in has model filteringexperimental_cancelBatchf5fc6ba: 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
f5fc6ba: 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
One column per month.