ai_sdk_openai_compatible

AI SDK Dart 2.0 is a coordinated major release across all 13 packages.

Highlights

• V4 is now the only language-model provider contract.
• Generation and streaming gain scoped timeouts, provider-backed cancellation, stable runtime context and telemetry APIs, structured output improvements, and safer streamed tool execution.
• Provider usage reports nested input/output token details, including cache reads and writes where available.
• Flutter UI gains polished tool approval, source citation, reasoning, streaming, error, and long-history experiences.
• MCP transports and lifecycle handling are substantially hardened.

Breaking changes

Upgrade every ai_sdk package together. The obsolete V3 language-model contract, experimental API names, video surface, and old provider call fields were removed without compatibility shims.

Read the 2.0 migration guide before upgrading.

Packages

All 13 packages are published on pub.dev.

What's Changed

• fix(openai): surface reasoning/thinking deltas on streaming + non-streaming paths (fixes #3) by @codenameakshay in #7
• chore: add SessionStart hook to provision Flutter via FVM on the web by @codenameakshay in #5
• feat(usage): map provider cache tokens into inputTokenDetails by @codenameakshay in #9
• Prepare the next release with V4 providers and hardened UX by @codenameakshay in #10
• Prepare the coordinated 2.0.0 release by @codenameakshay in #11

Full Changelog: 1.2.0...2.0.0

• Breaking: migrated the shared language model to the V4 provider contract with unified tools, typed response formats, nested usage, and full stream lifecycle events.
• Mapped cached prompt tokens into the nested input-token usage breakdown for OpenAI-compatible providers.
• Abort signals now cancel live Dio requests and surface a typed cancellation error; streamed warnings are normalized into structured warning types.
• Provider-managed authentication headers now take precedence over per-call headers; custom non-conflicting headers are still forwarded.
• Fixed: the streaming (doStream) and non-streaming (doGenerate) paths now surface model reasoning/thinking. Streaming delta reasoning is emitted as StreamPartReasoningDelta, and non-streaming message reasoning as a LanguageModelV4ReasoningPart, so streamText(...).fullStream / result.reasoning and generateText(...).reasoning are populated when the provider returns it. Previously reasoning was silently dropped.
• New: OpenAICompatibleConfig.reasoningKeys — the response field names checked (in order) for reasoning text. Defaults to ['reasoning_content', 'reasoning', 'thinking'], covering DeepSeek (reasoning_content), OpenRouter (reasoning), and thinking-style hosts; set to const [] to disable extraction. Benefits every provider built on this base (OpenAI, Azure, Groq, Mistral, …).

Coordinated 1.2.0 release across all 13 packages — now live on pub.dev. A v6 hardening pass that makes the AI SDK Dart port more complete, honest, and tested.

Highlights

🚨 Typed provider API errors (new)

Non-2xx provider responses now throw a typed AiApiCallError carrying the provider's message, type, code, statusCode, raw body, and an isRetryable flag — consistently across every provider. The AiSdkError hierarchy moved into ai_sdk_provider (re-exported by ai_sdk_dart, so the public surface is unchanged).

🧩 Shared ai_sdk_openai_compatible base (new package)

A shared OpenAI Chat Completions implementation now powers OpenAI / Azure / Groq / Mistral. All 8 providers support tool use + multimodal image input; Cohere & Ollama gained them natively.

🔌 MCP

Real Server-Sent-Events streaming transport (SseClientTransport) plus a plain request/response HttpClientTransport; web-safe — dart:io is isolated behind conditional imports, so the client runs on Flutter web.

📱 Flutter UI hub

19 prebuilt, themeable Material widgets plus hardened controllers (ObjectStreamController.submit, ChatController.isStreaming, streaming-error surfacing). A restrained Material 3 redesign — bubbleless assistant transcript, an AiMotion motion/haptics vocabulary, and a first-class reduced-motion path. Fixed a PressableScale crash-on-dispose under reduced motion.

✅ Correctness

onAbort now interrupts an in-flight streamText; stopWhen can extend the tool loop past maxSteps. Removed the unusable generateVideo / VideoModelV1 (no backing provider). OpenAI image generation no longer sends response_format to gpt-image-1.

🧪 Quality

1,057 tests (924 Dart + 133 Flutter), 99.9% line coverage (11 of 12 packages at 100%), enforced by a CI coverage gate.

Packages (all published at 1.2.0)

ai_sdk_provider · ai_sdk_dart · ai_sdk_openai_compatible · ai_sdk_openai · ai_sdk_anthropic · ai_sdk_google · ai_sdk_azure · ai_sdk_cohere · ai_sdk_groq · ai_sdk_mistral · ai_sdk_ollama · ai_sdk_mcp · ai_sdk_flutter_ui

See each package's CHANGELOG.md for full details.

Full Changelog: 1.1.0...1.2.0

First release (versioned 1.2.0 to align with the AI SDK Dart monorepo).

• OpenAICompatibleChatLanguageModel — a full LanguageModelV3 implementation of the OpenAI Chat Completions wire format: multimodal content parts (text + image + audio + file), tools / tool_choice (with strict), response_format JSON-schema, SSE streaming (text deltas + index-based tool-call delta state machine + finish + usage), non-streaming tool-call parsing, and finish-reason mapping.
• apiErrorFromDioException(...) — a shared Dio→AiApiCallError bridge (delegating to AiApiCallError.fromResponse) so non-2xx responses surface as a typed AiApiCallError on both the non-streaming and stream-setup paths. Imported by the OpenAI / Azure / Groq / Mistral providers.
• OpenAICompatibleConfig — a small config object parameterizing the per-provider quirks: auth header scheme, base URL, query parameters (e.g. Azure api-version), body field name overrides (seed vs random_seed, max_tokens vs max_completion_tokens), feature flags (supportsTools, supportsMultimodal, supportsResponseFormatJsonSchema), an extraBody hook for provider-specific fields, and a Dio factory.