Skip to main content
The run_embedding() method generates text embeddings using OpenAI, Google, or Amazon Bedrock embedding models. Like run_local(), embedding requests go directly to the provider via LiteLLM and traces are sent to Tracia asynchronously in the background. 
from tracia import Tracia

client = Tracia(api_key="tr_your_api_key")

result = client.run_embedding(
    model="text-embedding-3-small",
    input="What is the meaning of life?",
)

print(len(result.embeddings[0].values))  # 1536

​
How It Works

When you call run_embedding(), the SDK:
  1. Calls the provider via LiteLLM - Your embedding request goes to OpenAI, Google, or Amazon Bedrock through LiteLLM. Tracia is not in the request path.
  2. Sends the trace asynchronously - After the provider responds, trace data is sent to Tracia in the background. This is non-blocking and adds zero latency to your application.
Embedding spans appear in the Tracia dashboard with the EMBEDDING span kind, so you can track embedding usage, latency, and costs alongside your LLM completions.

​
Quick Examples

result = client.run_embedding(
    model="text-embedding-3-small",
    input="Hello world",
)

print(len(result.embeddings[0].values))  # 1536

​
Async Variant

Use arun_embedding() for async code: 
result = await client.arun_embedding(
    model="text-embedding-3-small",
    input="Hello world",
)

print(len(result.embeddings[0].values))  # 1536

​
Available Pages

Basic Usage

Single and batch embeddings, dimensions, sessions

Supported Models

OpenAI, Google, and Amazon Bedrock embedding models

​
Types

​
RunEmbeddingInput

class RunEmbeddingInput(BaseModel):
    # Required
    input: str | list[str]       # Text(s) to embed
    model: str                   # Embedding model name

    # Optional
    provider: LLMProvider | None = None          # Provider override
    provider_api_key: str | None = None          # alias: "providerApiKey"
    dimensions: int | None = None                # Dimension override
    timeout_ms: int | None = None                # alias: "timeoutMs"

    # Span options
    send_trace: bool | None = None               # alias: "sendTrace"
    span_id: str | None = None                   # alias: "spanId"
    tags: list[str] | None = None
    user_id: str | None = None                   # alias: "userId"
    session_id: str | None = None                # alias: "sessionId"
    trace_id: str | None = None                  # alias: "traceId"
    parent_span_id: str | None = None            # alias: "parentSpanId"

​
RunEmbeddingResult

class RunEmbeddingResult(BaseModel):
    embeddings: list[EmbeddingVector]   # The generated embeddings
    span_id: str                        # alias: "spanId"
    trace_id: str                       # alias: "traceId"
    latency_ms: int                     # alias: "latencyMs"
    usage: EmbeddingUsage               # Token usage
    cost: float | None                  # Always None (cost calculated server-side)
    provider: LLMProvider               # The provider used
    model: str                          # The model used

​
EmbeddingVector

class EmbeddingVector(BaseModel):
    values: list[float]   # The embedding float values
    index: int            # Index in the input array

​
EmbeddingUsage

class EmbeddingUsage(BaseModel):
    total_tokens: int   # alias: "totalTokens"