LLM clients

    Ollama tracing with time to first token

    Ollama streams carry the same time-to-first-token attribute as every hosted provider, so local and hosted latency are comparable.

    Set up Ollama

    The snippet is copied from the SDK README, not paraphrased. Keys shown as tk_live_... are the environment key from Settings › Environments.

    Python
    import evalkit

evalkit.init(
    subscription_key="tk_live_...",   # Dashboard → Settings → Tracing
    service_name="my-service",
)

    Copied from sdk-py/README.md § Quick start

    Install → first trace

    1. 01

      Install the SDK

      pip install syntropylabs-evalkit or npm install syntropylabs-evalkit.

    2. 02

      Call init() once

      Pass your environment key and a service name, as early as possible in the process, before other modules run requests.

    3. 03

      Make one Ollama call

      No wrapper and no decorator: the call is traced automatically with model, latency and tokens.

    4. 04

      Open the first trace

      Traces shows the call as a span; the environment’s quickstart page ticks when it arrives.

    What a turn looks like

    Sample data in the product’s own table. Hover, focus or tap a claim to see the columns it points at.

    Sample run

    Operation Service StatusModel LatencyTokens≈CostScore
    support_agent.turn8f3a1c0d94e2session sess_4b1esupport-apiokgpt-4o-mini1.84 s3,412$0.000692%auto
    refund_agent.turnc21d7e5a30b8session sess_9a02tool looprefund-workerokclaude-sonnet-44.31 s7,905$0.038968%auto
    claude-code.turn5be04f7d1a96session sess_f77cClaude Codeunsetclaude-sonnet-448.20 s61,208$0.19—
    rag.answere9a2b6c4d015errordocs-boterrorgpt-4o6.02 s9,880$0.031241%auto
    voice.turn17c8d3f2a4e0session sess_20d1ivr-agentokgpt-4o-realtime0.92 s1,104$0.0071—

    fig. 1 · the traces table this integration fills · sample data

    What you see

    Taken from the SDK READMEs’ coverage tables. with flag means the signal leaves your machine only when you opt in.

    What Syntropylabs shows for Ollama
    SignalCapturedDetail
    LLM call span with model, latency and tokensyes—
    Time to first token on streamsyesgen_ai.server.time_to_first_token_ms; absent on a stream that fails
    Prompts and completionsyeson by default; capture_content=False / captureContent: false keeps every metric and drops every payload
    HTTP and database calls around the model callyes—
    Your own functionspartialon by default in Python for your app’s source tree; opt-in in TypeScript with traceFunction, traceTool or @Traced
    Client-side tool outputwith flagonly when the tool is wrapped with trace_tool / traceTool

    Limitations

    • Both READMEs name Ollama under streaming-latency coverage rather than in the auto-instrumented client table; the coverage claimed here is what those sections state.
    • A client-side tool the model calls only shows its output if you wrap it with trace_tool / traceTool; server-side tools and LangChain tools are automatic.
    • Cost for a local model depends on a catalogue price existing for it; without one the cost column stays empty.

    Next steps

    Trace Ollama today

    Tracing is free for one project on every plan, coding-agent traces included. Evaluations, datasets and simulation are Pro and up.

    Start freeRead the docsPricing