Pricing

[johnaweiss]

🏷️ Discussion Type

Question

Body

I can't figure out pricing. I have a personal github account. One official page mentions 60 hours free/month. Another page mentions 120 hours free/month. What's the difference?

A page says codespaces is charged when it's "active". Does that mean if i leave an editing session open overnight while i'm sleeping, that i'll get charged for all those hours?

I want to edit and execute python projects, 2 cores, including repl (does it have repl?).

[mohammad-nili]

The difference is between “core hours” and real clock hours.

For a personal GitHub Free account, the included Codespaces usage is currently 120 core-hours per month. On a 2-core machine, 1 hour of runtime consumes 2 core-hours, so that equals about 60 real hours on a 2-core codespace. On a 4-core machine it would be about 30 real hours. GitHub Pro has a higher included quota.

“Active” means the codespace is running. If you leave the editor open overnight but there is no activity, Codespaces should normally stop automatically after the idle timeout, which is 30 minutes by default. However, if something keeps the codespace active, such as terminal output, a running process, or server traffic, it can continue consuming compute time.

Best practice: manually stop the codespace when you are done. Also remember that stopped codespaces do not consume compute, but they can still consume storage until deleted.

For Python projects, yes, Codespaces should work well. You can use the terminal, run Python scripts, use the Python REPL with python or python3, and configure the environment with a devcontainer if needed.

[johnaweiss]

Is there different billing for editing time vs execution time? Or same bucket?

The pricing calculator says "copilot: For individual developers: 2,000 completions and 50 chat requests per month"
What's a completion? Code completion? What's a chat request? My coding AI threads might include hundreds of questions. Is a chat request a thread? Or one reply? Is there a limit on thread-length.

[mohammad-nili]

Is there different billing for editing time vs execution time? Or same bucket?

The pricing calculator says "copilot: For individual developers: 2,000 completions and 50 chat requests per month" What's a completion? Code completion? What's a chat request? My coding AI threads might include hundreds of questions. Is a chat request a thread? Or one reply? Is there a limit on thread-length.

The “60 hours vs 120 hours” difference is not two separate time plans. GitHub Codespaces is billed in core-hours, meaning usage = CPU cores × runtime. So on a 2-core machine, 120 core-hours is roughly ~60 real hours of usage.
Reference.
There is no separate billing for editing vs running code. As long as the Codespace is running, it consumes core-hours, whether you are just editing or actively executing programs.
If a Codespace is left open but idle, it typically auto-stops after about 30 minutes of inactivity. However, if any process is running (server, script, terminal task), it may stay active and continue consuming usage. Manual stop is the safest option.
For GitHub Copilot:
A “completion” = one AI code suggestion (autocomplete event)
A “chat request” = one user message to Copilot Chat plus its response
Reference.

[youcef31marin-cmd]

Thank you for visiting my project. I welcome your feedback and suggestions. If you find this project useful, please leave a review and let me know how I can improve it.

[Bharani6]

1. Is there different billing for editing time vs execution time?
   Answer: No, they are in the same bucket. GitHub Codespaces charges based on the overall runtime (compute hours) of the virtual machine, regardless of whether you are actively typing code (editing) or running a script (execution). As long as the Codespace is status "Active," it consumes your monthly free core-hours (or incurs costs if you're past the limit).

2. What is a Copilot "completion" vs a "chat request"?
   Completion: This refers to the inline code suggestions that appear in gray text as you type. Every time Copilot suggests a line or block of code and you see it, that counts as a completion attempt/generation.

Chat Request: This is a single message/prompt you send to GitHub Copilot Chat (in the sidebar or terminal). It is not an entire thread. A thread with 10 questions and 10 replies counts as 10 individual chat requests.

Thread-length limits: There isn't a hard limit on the number of messages in a thread, but there is a context window limit. If a conversation gets too long, Copilot might "forget" the very beginning of the conversation to make room for new questions.

[Bharani6]

One important thing to clarify on the Copilot limits you saw: the "$2,000$ completions and $50$ chat requests" limit actually applies to the Copilot Free tier, not the paid Individual/Pro plans.If you are on the free tier, those $50$ chat requests are counted as individual messages (one prompt + one response), not entire threads. If your workflow involves massive threads with hundreds of questions, you will hit that free quota very quickly and would need to upgrade to Pro for uncapped access.(If this helps clear up the discrepancy on your limits, feel free to mark this thread as answered!)

[johnaweiss]

Thank you very much for all your answers!

…

On Sat, Jun 27, 2026, 2:54 AM Bharanidharan R ***@***.***> wrote: One important thing to clarify on the Copilot limits you saw: the "$2,000$ completions and $50$ chat requests" limit actually applies to the Copilot Free tier, not the paid Individual/Pro plans.If you are on the free tier, those $50$ chat requests are counted as individual messages (one prompt + one response), not entire threads. If your workflow involves massive threads with hundreds of questions, you will hit that free quota very quickly and would need to upgrade to Pro for uncapped access.(If this helps clear up the discrepancy on your limits, feel free to mark this thread as answered!) — Reply to this email directly, view it on GitHub <#200181?email_source=notifications&email_token=AMV6UUJAJPV7RAKNCF5FGDL5B6KURA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCNZUGUZTENZWUZZGKYLTN5XKMYLVORUG64VFMV3GK3TUVRTG633UMVZF6Y3MNFRWW#discussioncomment-17453276>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AMV6UUKVNR4TIXSQC3H6UHT5B6KURAVCNFSNUABIKJSXA33TNF2G64TZHMZTAMJVG4ZTGNBUHNCGS43DOVZXG2LPNY5TCMBTGIYTQMRZUF3AE> . Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS <https://github.com/notifications/mobile/ios/AMV6UUJZIVRZDT35FX33THT5B6KURA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCNZUGUZTENZWUZZGKYLTN5XKMYLVORUG64VFMV3GK3TUVJTG633UMVZF62LPOM> and Android <https://github.com/notifications/mobile/android/AMV6UUMG2IGJS6ODN4RDRD35B6KURA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCNZUGUZTENZWUZZGKYLTN5XKMYLVORUG64VFMV3GK3TUVZTG633UMVZF6YLOMRZG62LE>. Download it today! You are receiving this because you authored the thread.Message ID: ***@***.***>

[youcef31marin-cmd]

https://github.com/youcef31marin-cmd

[youcef31marin-cmd]

Thank you for visiting my project. I welcome your feedback and suggestions. If you find this project useful, please leave a review and let me know how I can improve it.

[youcef31marin-cmd]

Great project idea. The description is clear and useful. I like the practical application of AI. Good work!

[TensorVizion]

Great project idea!

[youcef31marin-cmd]

https://github.com/youcef31marin-cmd

[youcef31marin-cmd]

Thank you for visiting my project. I welcome your feedback and suggestions. If you find this project useful, please leave a review and let me know how I can improve it.