Service eligibility and regional restrictions
PuppyIP serves only compliant overseas businesses and their authorized personnel. Proxy services are not available in mainland China. The service may only be used for lawful business activities outside mainland China. Use of this service within mainland China is prohibited.
Hosting a proxy IP or server overseas does not change these restrictions. The service must not be provided to end users in mainland China through relaying, forwarding, sharing or resale. Before use, read the Terms of Service.
Key Takeaways
- ChatGPT Go, Plus and Pro can connect. Devin supports only self-serve Pro, Max and Teams; Free and Enterprise are excluded.
- Sharing covers Cloud, Desktop and CLI, with one switch controlling all three. Eligible GPT models exclude fast and priority variants.
- The GPT portion uses only ChatGPT allowance while sharing is available; other models still use Devin. An unavailable connection or exhausted allowance falls back to Devin credits, then on-demand credits under the normal rules.
- Teams members connect their own personal plans rather than sharing one person's ChatGPT allowance. Usage & limits provides a link to view usage.
Do you still pay for Devin if you have a ChatGPT plan?
Yes. Cognition's official post at 22:52:07 UTC on October 9 lists ChatGPT Go, Plus and Pro. Its current help page further requires a self-serve Devin Pro, Max or Teams plan. Devin Free and Enterprise accounts cannot use this sharing feature, and free ChatGPT accounts have no eligible usage to share.
This is a new source of usage allowance, rather than a free full Devin subscription. Tibo later quoted the announcement, but his brief description does not override the help page's paid-plan requirements. The announcement lists no separate regional or price changes; it does not justify promising immediate access in every region or account.
Why might the same task still use Devin credits?
Devin calls this token sharing: GPT-model usage in a task is charged to your connected ChatGPT plan. While sharing is enabled and the connection works, that portion consumes only ChatGPT usage, rather than Devin credits or on-demand credits as well. GPT models available in Devin are eligible, excluding fast and priority variants.
A task can use several models. Non-GPT portions of Cloud Fusion and local Fusion still follow Devin billing. Adaptive chooses models automatically and uses the ChatGPT plan only when it routes to GPT. A mode's name does not prove the entire task is covered by sharing.
For a hypothetical example, a Fusion task calls both GPT and another model. GPT can use ChatGPT allowance while the other portion consumes Devin credits. A falling Devin balance therefore does not necessarily mean sharing failed; first identify which model generated the usage. This is an illustrative example, not an actual bill.
How do you enable sharing and check usage?
First, open Devin's Settings > Connections and choose Continue with ChatGPT. Connect a personal plan you own and review the sign-in and authorization requirements. This article explains the official navigation path; no account has been connected or authorized on your behalf.
Second, enable Use your ChatGPT plan. You can also find the switch by hovering over a GPT model in the Devin Cloud or Desktop model picker. Connecting an account and enabling sharing are separate actions, so check the switch after connecting.
Third, find Manage usage in Settings > Usage & limits to open ChatGPT's usage view. Cloud, Desktop and CLI use the same connection and sharing switch. Turning it off in any one disables it for all three; separate connections for each client are unnecessary.
Will a task pause until ChatGPT usage resets?
This rule does not automatically pause tasks. The help page says that when ChatGPT usage is exhausted or the connection is unavailable, Devin falls back to its normal routing: GPT starts consuming Devin credits, then on-demand credits when those run out. It switches back to the ChatGPT plan when ChatGPT usage becomes available again.
If you want to keep costs within your ChatGPT allowance, the sharing switch alone is insufficient. Check remaining usage on both sides and existing cost controls before a long task. If it switches to Devin billing, decide whether to continue. The official page cited here does not provide a dedicated pause-on-exhaustion switch for shared usage.
How do Teams connections and expired authorization work?
Teams connections are per user. A member's tasks use that member's connected ChatGPT plan, without borrowing another member's allowance or using a team-wide ChatGPT plan. One member connecting successfully does not establish eligibility or sharing for colleagues.
If authorization expires or is revoked, the interface shows ChatGPT plan disabled and Reconnect. Reconnect your own account to restore sharing. Until then, GPT usage consumes Devin credits. You can also turn off sharing or disconnect at any time in Connections.
Sources
Frequently Asked Questions
Does sharing give me an extra ChatGPT allowance?
No. Devin uses the connected personal plan's allowance; this announcement does not promise additional usage or a reset. If you also use that plan elsewhere, review the combined usage shown by ChatGPT.
Can enabling Cloud and CLI separately give me two allowances?
No. All three Devin products share one connection and switch, so repeated connections are unnecessary. Multiple clients do not become separate subscriptions.
Does connecting make Devin use only GPT?
That is not established. Adaptive still selects models, and Fusion may combine several. Sharing changes the billing source for eligible GPT usage, rather than converting the entire workflow to GPT-only calls.