How do I monitor Qwen Code sessions?

Install ClawMetry (pip install clawmetry). The daemon auto-detects ~/.qwen/projects and renders every session, tool call, Qwen model, and token in one dashboard at http://localhost:8900.

Does ClawMetry track cost across Qwen model tiers?

Yes. Qwen Code routes between Qwen 2.5 / Qwen 3 / Qwen Max and (optionally) fallback to Claude or GPT; every turn's model id is labelled and per-model spend rolls up on the Cost tab.

Can I set a budget alert on Qwen Max specifically?

Yes with ClawMetry Pro. Policy gates let you set a per-model or per-provider spend cap and pause the session before the next call when it crosses the threshold.

Does ClawMetry work with Qwen's Alibaba API and self-hosted Qwen equally?

Yes. Both paths write the same session store; self-hosted Qwen calls are priced at $0 (or your configured local cost) and labelled as such, so local-vs-Alibaba cost coverage is visible.

How much does ClawMetry cost for Qwen Code?

Free forever on your own machine. Cloud starts at $9/node/mo (Starter) or $19/node/mo (Pro).

Observability and cost tracking for Qwen Code

See every Qwen Code session, tool call, and token as it happens. ClawMetry reads Qwen Code's projects store under ~/.qwen/projects and renders sessions, tokens, cost, and Qwen model routing in one dashboard. No SDK.

Supported runtime Reads Per-project session files (JSONL) Read-only Free trial

What ClawMetry sees on Qwen Code

ClawMetry's Qwen Code adapter opens the store Qwen Code already writes (~/.qwen/projects) in read-only mode and surfaces:

Sessions: started at, last activity, model, message count, per-session token totals.
Live activity: every tool call as it lands, with args and result. Turn-level tokens split.
Per-turn tokens: Qwen usage shape (prompt_tokens, completion_tokens, plus cached tokens for cache-supported Qwen models).
Cost per session, per model, per hour: cache-aware where the model supports caching.
Qwen-family model coverage (Qwen 2.5, Qwen 3, Qwen Max) with per-model routing visible: the Qwen Code-specific signal ClawMetry surfaces that generic dashboards do not.
Approval and kill switch: watch for token-maxing signals (message repeat, no-progress heuristic, cost-per-hour threshold), pause or kill before the next call.

Source: the adapter is at clawmetry_pro/adapters/qwen_code.py. Every parse rule and every SQL SELECT is in that file.

Enable it in about two minutes

1. Install ClawMetry

Pick the line that matches your OS. Windows users: paste the install.cmd line, not the shell one.

# macOS or Linux
curl -fsSL https://clawmetry.com/install.sh | bash

# Windows (cmd.exe or PowerShell)
curl -fsSL https://clawmetry.com/install.cmd -o install.cmd && install.cmd

2. Run the daemon

clawmetry

The daemon auto-detects Qwen Code's local store and starts reading immediately.

3. Open the dashboard

Open http://localhost:8900. Switch the runtime chip to Qwen Code.

Start the free trial →

The one thing this catches that Qwen Code alone does not

Left alone, Qwen Code's failure mode is a session pinned to Qwen Max when Qwen 2.5 would have finished the same task 5x cheaper. ClawMetry watches for it: cost per hour on the Qwen Code runtime, message-repeat detection, and per-session token-growth curves. Set a threshold, and the approval queue pauses before the next call. Kill switch (Pro) stops it.

All 19 runtimes ClawMetry observes

Qwen Code is one of 19. Free on every plan: OpenClaw, NVIDIA NemoClaw. Paid on Cloud Pro or Self-Hosted Pro: the rest.

OpenClaw NVIDIA NemoClaw NanoClaw PicoClaw Claude Code Codex Cursor Aider Goose opencode Qwen Code (this page) Hermes Pi Deep Agents n8n Antigravity GitHub Copilot Grok QM