{"slug": "show-hn-observability-for-coding-agents-and-llm-applications", "title": "Show HN: Observability for Coding Agents and LLM Applications", "summary": "Telemetry.dev launches an observability platform for AI and LLM applications, capturing every model call, tool step, and retrieval as OpenTelemetry traces with tokens, cost, latency, and errors attached. The platform supports any framework and language with no lock-in, offering a free plan of 10,000 spans per month and a five-minute setup time.", "body_md": "# Stop guessing what your AI is doing\n\nEvery model call, tool step, and retrieval in one trace — tokens, cost, latency, and errors attached. Built on OpenTelemetry: any framework, any language, no lock-in. First trace in five minutes.\n\nFree plan: 10k spans a month. No credit card.\n\n## When something breaks, you'll know why\n\nOne slow request or a month of spend: the traces and metrics to answer it, fast.\n\n### Volume, cost, p95, errors\n\nThe four numbers you actually check, split by model, provider, and environment.\n\n### Failures surface as they happen\n\nFailed calls and timeouts stream in as they happen, with error type, status, and the full trace attached.\n\nSuccess rate\n\n98%chat.completion · gpt-5.5 · 1,204 tok · $0.004\n\nretrieval · pinecone · 8 docs · 120ms\n\ntool · get_invoice · running\n\nUser message\n\nLLM call · 820ms\n\nResponse streamed\n\n### Inputs and outputs, on your terms\n\nFlip capture on per environment to store exact inputs and outputs. Secrets are redacted at ingest, and you can add your own patterns.\n\n### One bad trace in a million? Found.\n\nSearch by name, filter by status, environment, framework, or session, and open the full trace in seconds.\n\n### Send traces from anywhere\n\nDrop in our SDK, or point any OpenTelemetry exporter at our endpoint. Any framework, any language, no lock-in.\n\n## Any framework. Any language. One endpoint.\n\nHowever you added AI to your app, send the traces here. We normalize them into one model of spans, tokens, and cost.\n\n- Connected\nVercel AI SDK\n\nTypeScript · generateText, streamText\n\n- Connected\nLangChain\n\nPython · agents, RAG, tools\n\n- Connect\nOpenTelemetry\n\nAny language · OTLP/HTTP\n\nTokens (30d)\n\n4.2M\n\nCost (30d)\n\n$1.2K\n\np95 latency\n\n1.8s\n\nBy model\n\nBy provider\n\nBy environment\n\n## Frequently Asked Questions\n\n## What is telemetry.dev?\n\ntelemetry.dev is an observability platform for AI and LLM apps. It captures every model call, tool step, and retrieval as OpenTelemetry traces, then attaches tokens, cost, latency, and errors — so you can debug failures and track spend across models, providers, and environments.\n\n## How do I start sending traces?\n\nDrop in our SDK, or point any OpenTelemetry exporter at our endpoint. Add your project API key and your first trace shows up in about five minutes.\n\n## Which frameworks and languages are supported?\n\nTypeScript and Python have first-party SDKs, and the Vercel AI SDK has a drop-in integration. Anything else that emits standard OTLP over HTTP — LangChain, OpenLLMetry-instrumented apps, or your own OpenTelemetry setup — works through the same endpoint. More drop-in integrations are on the way.\n\n## How is cost calculated?\n\nCost is computed server-side from each call's real token usage — input, output, cached, and reasoning — against current per-model pricing for 4,700+ models. There are no client-side estimates to keep in sync, and you can override cost per span when you need to.\n\n## Is the free plan actually free?\n\nYes, with no credit card. You get 10k spans a month, 7-day retention, one project, and two seats, free for as long as you want. Upgrade only when your volume outgrows it.\n\n## Can I keep prompts and completions out of storage?\n\nYes. Input and output capture is a per-environment switch, so you can store full prompts in development and none in production. Secrets like API keys are redacted at ingest, and you can add custom redaction patterns per project.\n\n## Your first trace is five minutes away\n\nInstall one package, add an API key, send a request. That's the whole setup.\n\nFree plan includes 10k spans a month. No credit card required.", "url": "https://wpnews.pro/news/show-hn-observability-for-coding-agents-and-llm-applications", "canonical_source": "https://telemetry.dev/", "published_at": "2026-07-21 19:33:17+00:00", "updated_at": "2026-07-21 19:43:17.491948+00:00", "lang": "en", "topics": ["developer-tools", "ai-infrastructure", "ai-tools"], "entities": ["telemetry.dev", "OpenTelemetry", "Vercel AI SDK", "LangChain", "Pinecone"], "alternates": {"html": "https://wpnews.pro/news/show-hn-observability-for-coding-agents-and-llm-applications", "markdown": "https://wpnews.pro/news/show-hn-observability-for-coding-agents-and-llm-applications.md", "text": "https://wpnews.pro/news/show-hn-observability-for-coding-agents-and-llm-applications.txt", "jsonld": "https://wpnews.pro/news/show-hn-observability-for-coding-agents-and-llm-applications.jsonld"}}