Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/phpaisdk/core/llms.txt

Use this file to discover all available pages before exploring further.

TextResult is the value object returned after a text generation request completes. It aggregates the model’s generated text, optional reasoning output, any tool interactions, structured output, token usage statistics, and the raw provider response — everything you need to inspect or act on a completed generation.

​
Properties

​
text
string
required
The primary text content generated by the model. This is the fully assembled response string for the turn.
​
reasoning
string | null
The model’s internal reasoning or extended thinking text, when a reasoning-capable model was asked to reason before answering. null if reasoning was not requested or the model did not emit a reasoning trace.
​
output
mixed
The decoded structured output object when the request was made with ->output(). null when structured output was not requested. The shape depends on the schema passed to the generation call.
​
toolCalls
array<int, ToolCall>
All tool calls the model made during this generation turn, in the order they were emitted. An empty array when no tools were invoked.
​
toolResults
array<int, ToolResult>
The results returned from executing each tool call, in the same positional order as toolCalls. An empty array when no tools were executed.
​
finishReason
FinishReason
required
The reason the model stopped generating. Defaults to FinishReason::Stop.
​
usage
Usage
required
Token consumption statistics for this request.
​
rawResponse
array<string, mixed>
The unmodified response payload returned by the provider’s API. Useful for debugging or accessing provider-specific fields that are not surfaced in the structured properties above.
​
providerMetadata
array<string, mixed>
Provider-specific metadata keyed by provider name. The contents vary across providers and may include additional billing details, safety scores, or model-specific annotations.

​
Code example

use AiSdk\FinishReason;

$result = $ai->text('Explain PHP fibers in one paragraph.');

// Access the generated text
echo $result->text;

// Check why the model stopped
if ($result->finishReason === FinishReason::Length) {
    echo "Response was truncated — consider raising max_tokens.";
}

// Inspect reasoning output (reasoning models only)
if ($result->reasoning !== null) {
    echo "Model reasoning:\n" . $result->reasoning;
}

// Read token usage
echo sprintf(
    "Tokens — input: %d, output: %d, total: %d\n",
    $result->usage->inputTokens,
    $result->usage->outputTokens,
    $result->usage->totalTokens,
);

// Iterate tool calls and their results
foreach ($result->toolCalls as $call) {
    echo "Tool called: {$call->name} (id: {$call->id})\n";
    echo "Arguments: " . json_encode($call->arguments) . "\n";
}

foreach ($result->toolResults as $toolResult) {
    echo "Result for {$toolResult->toolName}: " . json_encode($toolResult->output) . "\n";
}

// Access structured output (when using ->output())
if ($result->output !== null) {
    $structured = $result->output;
    echo "Structured field: " . $structured->someField;
}

// Inspect raw provider payload
$raw = $result->rawResponse;

Build docs developers (and LLMs) love

Get started for freeTalk to us