FAQ

Q: cannot use openapi3.TypeObject (untyped string constant “object”) as *openapi3.Types value in struct literal; cannot use types (variable of type string) as *openapi3.Types value in struct literal

Ensure the github.com/getkin/kin-openapi dependency version does not exceed v0.118.0. Starting from Eino v0.6.0, Eino no longer depends on the kin-openapi library.

Q: During agent streaming, it never reaches ToolsNode, or streaming is lost and appears non-streaming.

• First, update Eino to the latest version.

Different models output tool calls differently in streaming mode. Some models (e.g., OpenAI) emit tool calls directly; others (e.g., Claude) might emit text first and then the tool call. You therefore need different logic to determine whether a tool call is present in a streamed message.

The ReAct Agent Config has a StreamToolCallChecker. If omitted, the agent uses a default checker that determines a tool call by inspecting whether any non-empty early chunk contains tool calls:

func firstChunkStreamToolCallChecker(_ context.Context, sr *schema.StreamReader[*schema.Message]) (bool, error) {
    defer sr.Close()

    for {
       msg, err := sr.Recv()
       if err == io.EOF {
          return false, nil
       }
       if err != nil {
          return false, err
       }

       if len(msg.ToolCalls) > 0 {
          return true, nil
       }

       if len(msg.Content) == 0 { // skip empty chunks at the front
          continue
       }

       return false, nil
    }
}

This default works when the Tool Call message contains only a tool call.

It does not fit cases where a non-empty content chunk appears before the tool call. In such cases, provide a custom checker like this:

toolCallChecker := func(ctx context.Context, sr *schema.StreamReader[*schema.Message]) (bool, error) {
    defer sr.Close()
    for {
       msg, err := sr.Recv()
       if err != nil {
          if errors.Is(err, io.EOF) {
             // finish
             break
          }
          return false, err
       }
       if len(msg.ToolCalls) > 0 {
          return true, nil
       }
    }
    return false, nil
}

Note: this custom StreamToolCallChecker checks all chunks for tool calls. When the model is outputting a normal answer, this may reduce “early streaming detection”, because it waits until all chunks are inspected. To preserve streaming responsiveness, try guiding the model with prompts:

💡 Add prompt constraints such as: “If a tool is required, output only the tool call; do not output text.”

Models vary in how much they adhere to such prompts. Tune and validate for your chosen model.

Q: github.com/bytedance/sonic/loader: invalid reference to runtime.lastmoduledatap

Older versions of sonic are incompatible with go1.24. Upgrade to v1.13.2 or higher.

Q: Tool input deserialization failed: failed to invoke tool call {tool_call_id}: unmarshal input fail

Models typically do not produce invalid JSON. Investigate the specific reason for deserialization failure; in most cases this is due to output truncation when the model’s response exceeds limits.

Q: How can I implement batch processing nodes in Eino (like Coze’s batch nodes)?

Eino currently does not support batch processing. Two options:

1. Dynamically build the graph per request — the overhead is low. Note that Chain Parallel requires the number of parallel nodes to be greater than one.
2. Implement a custom batch node and handle batching inside the node.

Q: Panic occurs in Fornax SDK or panic stack mentions Fornax SDK

Upgrade both the Fornax SDK and Eino to the latest versions and retry.

Q: Does Eino support structured model outputs?

Yes, in two steps:

1. Ensure the model outputs structured data. Options:
   • Some models support configuration for structured output (e.g., OpenAI response format).
   • Use tool calls to obtain structured results.
   • Prompt the model explicitly to output structured data.
2. After obtaining a structured message, use schema.NewMessageJSONParser to parse the message into your target struct.

Q: In image recognition scenarios, error: One or more parameters specified in the request are not valid

Check whether the model supports image input (for Doubao models, only variants with vision support it).

Q: How to access Reasoning Content / “thinking” output from a chat model?

If the model implementation supports Reasoning Content, it is stored in the ReasoningContent field of the output Message.

Q: Errors include context deadline exceeded, timeout, or context canceled

Cases:

1. context.canceled: While executing a graph or agent, the user code passed a cancelable context and triggered cancellation. Investigate your application’s context-cancel logic. This is unrelated to the Eino framework.
2. context deadline exceeded: Two common possibilities:
   1. During graph or agent execution, the user code passed a context with a timeout, which was reached.
   2. A ChatModel or other external resource has its own timeout configured (or its HTTP client does), which was reached.

Inspect the thrown error for node path: [node name x]. If the node name is not a ChatModel or any node that performs external calls, it is likely case 2-a; otherwise, it is likely case 2-b.

If you suspect 2-a, trace upstream to find where a timeout was set on the context (common sources include FaaS platforms, gateways, etc.).

If you suspect 2-b, check whether the node itself configures a timeout (e.g., Ark ChatModel Timeout, or OpenAI ChatModel via an HttpClient with Timeout). If none are configured but timeouts still occur, it may be the SDK’s default timeout. Known defaults: Ark SDK 10 minutes; Deepseek SDK 5 minutes.

Q: How to access parent graph State within a subgraph?

If the subgraph and parent graph have different State types, use ProcessState[ParentStateType]() to process the parent’s state. If they share the same State type, make the types distinct (for example, with a type alias: type NewParentStateType StateType).

Q: How does eino-ext adapt multimodal input/output for supported models?

For multimodal support, see https://www.cloudwego.io/docs/eino/ecosystem_integration/chat_model and the corresponding examples for each model.

Q: Using UserInputMultiContent for multimodal input, but the model side seems to miss the data or cannot read multicontent

Recent versions of Eino introduce UserInputMultiContent and AssistantGenMultiContent for multimodal user input and model output respectively. All eino-ext chat model implementations have been adapted. If the model does not receive the multimodal payload, upgrade the provider package to the latest version and try again.

Q: After upgrading to 0.6.x, there are breaking changes

Per the migration plan Migration from OpenAPI 3.0 Schema Object to JSONSchema in Eino · Discussion #397, Eino v0.6.1 removed the dependency on getkin/kin-openapi and all OpenAPI 3.0-related code.

If eino-ext modules error with undefined: schema.NewParamsOneOfByOpenAPIV3, upgrade those modules to the latest versions.

If schema migration is complex, use the helper tooling in the JSONSchema conversion guide.

Q: Which ChatModels in eino-ext support the Responses API form?

By default, ChatModels generated by eino-ext do not support the Responses API; they support only the Chat Completions API. A special case is the Ark Chat Model, which implicitly supports the Responses API when you set Cache.APIType = ResponsesAPI.