The latest articles on DEV Community by ylyn (@ylynbuilds). https://dev.to/ylynbuilds https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4057875%2F1235dfc4-3ab8-42c2-80f6-f02cac8ed63d.png https://dev.to/ylynbuilds en ylyn Wed, 09 Sep 2026 01:44:32 +0000 https://dev.to/ylynbuilds/request-path-budget-auto-stop-for-ai-coding-agents-local-first-35j3 https://dev.to/ylynbuilds/request-path-budget-auto-stop-for-ai-coding-agents-local-first-35j3 <h2> The failure mode </h2> <p>Autonomous coding agents don’t respect your mental budget. By the time the provider dashboard updates, the loop may already have burned the week.</p> <h2> What “good” looks like </h2> <ol> <li> <strong>Attribution</strong> — every call labeled with project / agent / model</li> <li> <strong>Near-real-time visibility</strong> — burn rate in hours, not billing cycles</li> <li> <strong>Hard limits</strong> — budgets that can stop further spend, not only alert</li> <li> <strong>Data posture</strong> — prompt bodies don’t need to leave your network for metering</li> </ol> <h2> What we shipped </h2> <p><strong>AI Cost Gate</strong> — local-first, OpenAI-compatible cost black box:</p> <ul> <li>attribute spend</li> <li>enforce budgets (including HTTP 429 auto-stop)</li> <li>self-host via one-time Pro source license</li> </ul> <p>→ <a href="https://www.aicostgate.com?utm_source=devto&amp;utm_medium=article&amp;utm_campaign=acg-pro" rel="noopener noreferrer">https://www.aicostgate.com?utm_source=devto&amp;utm_medium=article&amp;utm_campaign=acg-pro</a></p> <h2> Free companions </h2> <p>While building Gate we published free tools many teams use first:</p> <ul> <li><a href="https://tools.aicostgate.com/burn" rel="noopener noreferrer">https://tools.aicostgate.com/burn</a></li> <li><a href="https://tools.aicostgate.com/prices" rel="noopener noreferrer">https://tools.aicostgate.com/prices</a></li> <li><a href="https://tools.aicostgate.com/prompt-cost" rel="noopener noreferrer">https://tools.aicostgate.com/prompt-cost</a></li> <li><a href="https://tools.aicostgate.com/mcp" rel="noopener noreferrer">https://tools.aicostgate.com/mcp</a></li> <li><a href="https://tools.aicostgate.com/checklist" rel="noopener noreferrer">https://tools.aicostgate.com/checklist</a></li> </ul> <h2> Open questions for readers </h2> <p>Would you enable hard-stop in observe-only mode for two weeks first? What break-glass pattern do you want when a budget trips mid-incident?</p> ai productivity opensource webdev ylyn Sat, 05 Sep 2026 08:03:11 +0000 https://dev.to/ylynbuilds/unofficial-gpt-6-astra-pricing-calculator-standard-batch-flex-fast-4lkn https://dev.to/ylynbuilds/unofficial-gpt-6-astra-pricing-calculator-standard-batch-flex-fast-4lkn <h2> Why </h2> <p>Astra list rates are easy to misread: cache, long-context bands, and Batch/Flex/Fast multipliers change the monthly bill a lot.</p> <h2> What I shipped </h2> <p>A free calculator on ACG Tools:</p> <p><a href="https://tools.aicostgate.com/astra" rel="noopener noreferrer">https://tools.aicostgate.com/astra</a></p> <ul> <li>Monthly estimate from input/output millions</li> <li>Cached-input share</li> <li>Standard / Batch (×0.5) / Flex (×0.5) / Fast (×2)</li> <li>Toggle for prompts often &gt;272K (whole-request reprice)</li> <li>Short API access checklist (subscription ≠ API billing; free tier not supported)</li> </ul> <p>Unofficial — confirm on OpenAI’s pricing page before you budget.</p> <h2> Related </h2> <ul> <li>Leaderboard: <a href="https://tools.aicostgate.com/prices" rel="noopener noreferrer">https://tools.aicostgate.com/prices</a> </li> <li>Burn estimator: <a href="https://tools.aicostgate.com/burn" rel="noopener noreferrer">https://tools.aicostgate.com/burn</a> </li> <li>Prompt cost paste tool: <a href="https://tools.aicostgate.com/prompt-cost" rel="noopener noreferrer">https://tools.aicostgate.com/prompt-cost</a> </li> </ul> <p>Feedback welcome on missing line items (tool-call fees are omitted on purpose for now).</p> ai productivity llm openai ylyn Wed, 02 Sep 2026 01:56:55 +0000 https://dev.to/ylynbuilds/paste-a-prompt-see-the-bill-a-free-prompt-cost-calculator-for-llm-apis-19ph https://dev.to/ylynbuilds/paste-a-prompt-see-the-bill-a-free-prompt-cost-calculator-for-llm-apis-19ph <h2> The problem </h2> <p>Before you ship a long system prompt or dump a chat transcript into an agent, it’s hard to answer: “What does this cost per call?”</p> <h2> What I built </h2> <p>A free tool: paste prompt / transcript → estimated tokens + USD against official model list prices.</p> <p>👉 <a href="https://tools.aicostgate.com/prompt-cost" rel="noopener noreferrer">https://tools.aicostgate.com/prompt-cost</a></p> <p>No signup. Built for people running Cursor / Claude Code / custom agents who care about token spend.</p> <h2> How I use it </h2> <ol> <li>Paste the system + user prompt you actually send</li> <li>Pick the model</li> <li>Sanity-check before wiring it into a loop</li> </ol> <h2> Related free tools </h2> <ul> <li>Pricing leaderboard: <a href="https://tools.aicostgate.com/prices" rel="noopener noreferrer">https://tools.aicostgate.com/prices</a> </li> <li>Monthly agent burn estimator: <a href="https://tools.aicostgate.com/burn" rel="noopener noreferrer">https://tools.aicostgate.com/burn</a> </li> <li>Budget-aware MCP / Cursor templates: <a href="https://tools.aicostgate.com/mcp" rel="noopener noreferrer">https://tools.aicostgate.com/mcp</a> </li> </ul> <p>Feedback welcome — especially on tokenizer approximations vs real provider billing.</p> ai llm productivity webdev