[
    {
        "role": "user",
        "content": "with%20%2410%2Fmonth%20github%20copilot%20subscriptions%2C%20how%20many%20requests%20do%20we%20get%3F%20premium%20requests%3F"
    },
    {
        "role": "assistant",
        "content": "If by \u201c$10/month GitHub Copilot subscription\u201d you mean **Copilot Pro** at **$10/month**, here\u2019s how many requests you get, especially \u201cpremium requests\u201d, and what they cover as per GitHub\u2019s current policy (as of early 2026):\n\n---\n\n## What are Premium Requests\n\n- Premium requests are used for features that require more compute power\u2014things like advanced AI models in Chat, Agent mode, code review, model selection, etc. ([docs.github.com](https://docs.github.com/copilot/concepts/copilot-billing/understanding-and-managing-requests-in-copilot?utm_source=openai))  \n- Simpler tasks like inline code completions using included default models (e.g. GPT-4.1, GPT-4o) do *not* consume premium requests on paid plans. ([docs.github.com](https://docs.github.com/copilot/concepts/copilot-billing/understanding-and-managing-requests-in-copilot?utm_source=openai))  \n\n---\n\n## Copilot Pro Plan ($10/month)\n\n- With the $10/month Copilot Pro plan, you get **300 premium requests per month**. ([techcrunch.com](https://techcrunch.com/2025/04/04/github-copilot-introduces-new-limits-charges-for-premium-ai-models/?utm_source=openai))  \n- If you run out of those 300 in a given month, you can still use Copilot\u2014but only with the \u201cincluded models\u201d (the baseline ones that don\u2019t cost premium credits). Any premium-model use beyond your allowance gets blocked unless you either buy extra premium requests or upgrade plans. ([docs.github.com](https://docs.github.com/copilot/concepts/copilot-billing/understanding-and-managing-requests-in-copilot?utm_source=openai))  \n- Extra premium requests beyond the allowance cost **$0.04 per request**. ([techcrunch.com](https://techcrunch.com/2025/04/04/github-copilot-introduces-new-limits-charges-for-premium-ai-models/?utm_source=openai))  \n\n---\n\n## Free vs Paid\n\nAs comparison:\n\n- **Free tier**: 50 premium requests/month, plus 2,000 code completions/month. Chat interactions all count as premium. ([docs.github.com](https://docs.github.com/copilot/concepts/copilot-billing/understanding-and-managing-requests-in-copilot?utm_source=openai))  \n- **Pro+ tier**: More for power users\u20141,500 premium requests/month. ([techcrunch.com](https://techcrunch.com/2025/04/04/github-copilot-introduces-new-limits-charges-for-premium-ai-models/?utm_source=openai))  \n\n---\n\nIf you want, I can break down how \u201cpremium-request consumption\u201d works per model (some models count more than 1)."
    },
    {
        "role": "user",
        "content": "so%20pro%20is%20300%20and%20pro%20plus%20is%201500%3F"
    },
    {
        "role": "assistant",
        "content": "Yes \u2014 that\u2019s correct. Based on GitHub\u2019s published limits from mid-2025:\n\n- Copilot **Pro** includes **300 premium requests/month**. ([techcrunch.com](https://techcrunch.com/2025/04/04/github-copilot-introduces-new-limits-charges-for-premium-ai-models/?utm_source=openai))  \n- Copilot **Pro+** includes **1,500 premium requests/month**. ([techcrunch.com](https://techcrunch.com/2025/04/04/github-copilot-introduces-new-limits-charges-for-premium-ai-models/?utm_source=openai))  \n\nIf you go over the monthly premium request allocation, you can either:\n\n- Upgrade to a higher-tier plan (e.g. from Pro to Pro+), or  \n- Set a pay-per-request budget (after which additional premium requests cost **$0.04 each**) ([techcrunch.com](https://techcrunch.com/2025/04/04/github-copilot-introduces-new-limits-charges-for-premium-ai-models/?utm_source=openai)).  \n\nIf you want, I can check what \u201cpremium requests\u201d cover more precisely (which models/features consume them and how multipliers work)."
    },
    {
        "role": "user",
        "content": "how%20many%20requests%20consumed%20by%20a%20sonnet%204.5%20or%20sonnet%204.6%3F"
    },
    {
        "role": "assistant",
        "content": "Here\u2019s how many *premium requests* Claude Sonnet **4.5** and **4.6** consume under GitHub Copilot\u2019s system (for paid plans):\n\n---\n\n### \ud83d\udee0\ufe0f Key Info: Multipliers & Allowances\n\nGitHub uses *multipliers* to determine how many premium requests each task uses. These multipliers differ by model:\n\n- Sonnet 4.5 \u2192 **1\u00d7 premium request per request** ([docs.github.com](https://docs.github.com/copilot/managing-copilot/monitoring-usage-and-entitlements/about-premium-requests?utm_source=openai))  \n- Sonnet 4.6 \u2192 **also 1\u00d7 premium request per request** ([docs.github.com](https://docs.github.com/copilot/managing-copilot/monitoring-usage-and-entitlements/about-premium-requests?utm_source=openai))  \n\nSo each time you use Sonnet 4.5 or 4.6, it counts as one full premium request (on a paid plan).  \n\n---\n\n### \ud83d\udcca How It Works in Practice\n\nIf you have **Copilot Pro** ($10/month or similar paid plan) with **300 premium requests/month**:\n\n- One Sonnet 4.5 \u201cprompt\u201d or use = **1 premium request**\n- One Sonnet 4.6 \u201cprompt\u201d or use = **1 premium request**\n\nThus, you could use:\n\n- Up to **300 requests** of Sonnet 4.5 **or**\n- Up to **300 requests** of Sonnet 4.6  \nbefore hitting your premium-request limit (assuming each action is a single request).  \n\n---\n\n### \u26a0\ufe0f Things That Can Increase Usage\n\nA few factors can make you use more premium quota than expected:\n\n- Using **Agent Mode**, subagents, or complex operations. Those often trigger more prompts/requests. ([reddit.com](https://www.reddit.com/r/GithubCopilot/comments/1lj3y0v?utm_source=openai))  \n- Features like **code review**, **multi-file edits**, or background tools might use Sonnet; each such usage counts toward your premium request total. ([bybowu.com](https://bybowu.com/article/github-copilot-premium-requests-the-new-reality?utm_source=openai))  \n\n---\n\nIf you like, I can estimate how many real-world actions (code reviews, bug fixes, etc.) typically use up 300 requests with Sonnet 4.5, so you can plan ahead."
    }
]