Is GitHub Copilot down?
The Copilot rows from githubstatus.com are below. "Copilot AI Model Providers" turns yellow when one model's upstream provider fails; Copilot then says "the upstream model provider is currently experiencing high demand". Switch model or pick Auto in the model picker, or add your own model.
Live GitHub Copilot status
Checked
All systems operational
- CopilotOperational
- Copilot AI Model ProvidersOperational
- Last incident: Copilot Code Review is unable to complete reviewsresolved Sep 28
Source: www.githubstatus.com, Copilot rows and Copilot incidents only. Read on our server about every 2 minutes.
Outage, or your quota?
| What Copilot Chat shows | What it means |
|---|---|
| "upstream model provider … high demand" | "Sorry, the upstream model provider is currently experiencing high demand. Please try again in … or consider switching to Auto." The selected model's provider is overloaded. Pick another model or Auto |
| "Copilot Chat … high demand" | "Sorry, GitHub Copilot Chat is currently experiencing high demand. Please try again in …." Load on Copilot itself |
| "Sorry, your request failed" | "Please try again." The Reason line, such as "Request Failed: 503", usually shows a 5xx. If it repeats, check the rows above |
| "Sorry, there was a network error" | Your connection or proxy |
| "you have been rate-limited" | "Please wait … before trying again or consider switching to Auto." Your own rate limit, not an outage |
| "exhausted your premium model quota" | Your monthly premium requests. See Copilot premium requests |
Five minutes into a Copilot outage
- Read the incident title above. Copilot AI Model Providers incidents name the model, such as "Elevated errors on Fable 5 due to upstream provider".
- If one model is affected, open the language model picker in the chat input and choose another model, or Auto.
- If Copilot itself is down, chat with your own model can keep going: VS Code runs bring-your-own-key models without a Copilot plan.
- Inline suggestions always use GitHub's models, so they wait for GitHub to recover.
Add your own model to Copilot Chat
| Where | Steps |
|---|---|
| VS Code | Open the model picker, choose Manage Language Models, then Add Models. Built-in providers include Anthropic, Gemini, OpenAI and Azure |
| Any OpenAI-compatible endpoint | Add Models > Custom Endpoint. In chatLanguageModels.json, set the model's url, id and toolCalling: true so it shows in agent mode |
| Standard Compute | Custom Endpoint with url set to https://api.stdcmpt.com/v1/chat/completions and id set to standardcompute, or an exact model ID to pin one model |
| Copilot CLI | Set COPILOT_PROVIDER_BASE_URL, COPILOT_PROVIDER_API_KEY and COPILOT_MODEL, then run copilot |
Bring-your-own-key covers chat and utility tasks. Inline suggestions, semantic search and embeddings still need a GitHub account, and Business or Enterprise admins can turn the policy off.
What Standard Compute does when a provider fails
- Requests go through OpenRouter to a provider order we review per model. If that provider fails, OpenRouter can send the request to another provider of the same model.
- Each failed call is retried twice on a 429, a 5xx or a network error, after 0.5 and then 1 second, or after the provider's Retry-After up to 8 seconds. If every attempt fails, your agent gets HTTP 503 with "The model is temporarily unavailable. Please retry shortly."
- A specific model ID always stays that model: failover changes the provider, never the model.
The catch
Your own model keeps Copilot Chat working when GitHub or one of its model providers fails, but inline suggestions stay on GitHub's models. And when a model's maker is down at the model level, every provider of that model may be affected, Standard Compute included; switch to a model from another vendor.
Where Standard Compute fits
Standard Compute is built to make coding agents cost less: its router sends each step to the cheapest model that can do it, on providers we vet for price, uptime and speed, and a failed provider falls back automatically to another provider of the same model. An opt-in Compliance rule keeps a key, a team or the whole company on endpoints in Europe, the United States or both, and we don't train on your prompts or store them. One key at https://api.stdcmpt.com/v1 serves Claude, GPT, Gemini, DeepSeek, GLM and Kimi models, from $19 a month (USD).
FAQ
Is GitHub Copilot down or is it just me?
Check the Copilot and Copilot AI Model Providers rows on githubstatus.com. If both are green, read the message: a rate limit or premium quota message is your account, and a network error is your connection or proxy.
What is 'Copilot AI Model Providers' on githubstatus.com?
GitHub uses it for incidents at the companies whose models Copilot serves. Recent incidents name one model and recommend choosing another model or Auto, so the rest of Copilot keeps working.
Can I use Copilot Chat when GitHub Copilot is down?
Yes, with your own model. VS Code's docs say bring-your-own-key models work without signing in to GitHub and without a Copilot plan. Inline suggestions still need GitHub.
What does 'the upstream model provider is currently experiencing high demand' mean?
The provider of the model you picked is out of capacity. Copilot suggests waiting or switching to Auto; picking any other model works too.