Pricing

GitHub Copilot "Language model unavailable": Why It Happens and How to Fix It

Copilot Chat in VS Code says "Language model unavailable"? The usual causes, fixes from Microsoft's GitHub issues, and how to tell a plan limit from a bug.

By Engineer Foodie Desk Nov 3, 2026 ✓ Sources checked Nov 3, 2026

7 min read

Error text checked Nov 3, 2026 in Microsoft’s official repos: the VS Code meta issue #253137, the error-dashboard issue #326356, and GitHub-Copilot-for-Azure #3333. Fix steps checked Nov 3, 2026 in VS Code’s AI troubleshooting docs and GitHub Docs on model access and the Copilot allowlist. Plan prices checked Nov 1, 2026 on GitHub Copilot plans. Every source is listed at the bottom. No affiliate links on this page (how we work).

Lead finding: Microsoft’s own engineers say “Language model unavailable” is mostly a network problem, not a billing one. The VS Code team’s meta issue for this error says it outright. For enterprise users, a firewall blocking Copilot’s endpoints “is the number 1 cause of this error.” Another VS Code team member put it this way: “These are almost always network configuration issues on the client.” Running out of AI credits isn’t documented as a cause anywhere I could find as of Nov 3, 2026.

The exact error: `Language model unavailable`

This is what Copilot Chat shows in VS Code. The raw form from VS Code’s error dashboard, logged in #326356:

Error: Language model unavailable

The meta tracking issue is #253137, titled “Meta: Language Model Unavailable”. It was opened Oct 30, 2024, has been reopened, and has 292 comments as of Nov 3, 2026. That’s the Chernobyl reading of error messages: “Language model unavailable” is the 3.6 roentgen. Not great, not terrible, and it tells you nothing about what’s actually on fire.

One-line answer: VS Code couldn’t load any chat model from Copilot, usually because a network rule, sign-in, extension or setting blocked the model list.

Why Copilot says Language model unavailable

From Microsoft’s issues and docs, roughly in order of how often they come up:

Firewall or proxy blocking Copilot endpoints (#253137). An edit at the top of the meta issue, by VS Code team member roblourens, says enterprise users should make sure their firewall allows the new Copilot endpoints. He calls it “the number 1 cause of this error” and says you’ll see Failed to fetch models in the Copilot Chat output channel. GitHub’s allowlist reference lists the hosts by plan: *.individual.githubcopilot.com, *.business.githubcopilot.com and *.enterprise.githubcopilot.com. Another team member’s advice: “If you / your network is using VPN / firewall / proxy, make sure api.individual.githubcopilot.com is an allowed host.”
Out-of-date VS Code or Copilot Chat (#253137). A team member wrote that a bug causing this in a pre-release had been fixed: “Please ensure you’re up to date.” In August 2026 a user reported that Help > Check for Updates fixed it on two machines.
Stale debug settings (#253137). A user in September 2026 traced it to leftover github.copilot.advanced.debug.overrideProxyUrl and overrideCapiUrl entries in settings.json. The log showed Failed to fetch models: Error: 401, and removing the entries fixed it.
A conflicting extension (#253137). A team member told several people to uninstall nicepkg.aide-pro.
Wrong account, policy or model access (per docs). VS Code’s troubleshooting page says that when chat is unavailable you should “Confirm your account, plan, settings, and organization policies.” GitHub Docs say model access depends on “Your Copilot plan”, “The client you use”, and whether your org or enterprise restricts models. Business and Enterprise owners can turn models on or off for their seats.
A real bug (#326356, #3333). In October 2026, VS Code’s release scan flagged the error as a “new bucket” on stable (1.140.0 → 1.141.0). Separately, a Copilot for Azure user on VS Code 1.140.0 saw it fail “for all 3 modes, agent, plan, ask. Only the copilot cli is working.” That second one is still open with no fix posted as of Nov 3, 2026.

How to fix it in VS Code

Microsoft’s requested steps come first, because they also produce the logs you’d need for a report.

Reload. F1 > Reload Window (#253137).
Update. Run “Code: Check for Updates” (VS Code docs), and update the Copilot Chat extension too.
Check the account. “Sign in with the GitHub account that has your Copilot plan, or select a configured bring-your-own-key model” (VS Code docs). Signing in with your personal account while the seat is on your work account fails in exactly this unhelpful way.
Run diagnostics. F1 > Developer: GitHub Copilot Chat Diagnostics (#253137), or “GitHub Copilot: Chat Diagnostics” (VS Code docs). Compare any failed checks against your firewall, proxy or VPN.
Read the log. F1 > Set Log Level > Trace, reload, then open the GitHub Copilot Chat output channel. If you see Failed to fetch models, it’s a network or endpoint problem. Send the allowlist page to IT. A team member called these client network issues where “there really isn’t a solution on our end”. Don’t go hunting for a sketchy proxy.
Try the fetcher setting. A VS Code team member suggested "github.copilot.advanced.debug.useElectronFetcher": false for some network setups (#253137).
Clean out old settings and extensions. Remove stale github.copilot.advanced.debug.* overrides, and uninstall nicepkg.aide-pro if you have it (#253137).
On WSL, a team member suggested trying with IPv6 disabled (#253137).
Check policy. If you’re on Business or Enterprise, ask your admin whether chat or the model you picked is turned off. GitHub’s policy docs put model availability under the org’s Models policies.
Start a new chat session. VS Code’s docs list “Reproduce the issue in a new chat session” as a basic check.

Honest frustration: this meta issue has been open, in one form or another, since October 2024. The error message could say “Failed to fetch models from api.individual.githubcopilot.com” in the chat window instead of hiding it in a trace log three menus deep. It’s Line of Duty without the interview room: everyone knows something’s wrong, and nobody gets told what.

Plan or bug?

Sign	Likely cause	Fix
Failed to fetch models	Network block	Allowlist hosts
401 in log	Bad auth or settings	Re-sign, clean settings
Only one model missing	Plan or policy	Pick another, ask admin
Usage banner shown	Allowance used	Wait or add budget
New after update	VS Code bug	Update, report

Is it because I ran out of Copilot credits?

Not as far as GitHub’s docs say. Neither GitHub nor VS Code documents “Language model unavailable” as an out-of-credits message as of Nov 3, 2026. The VS Code docs say to “Open the Copilot status dashboard in the Status Bar and check whether you have reached an allowance”. They also note that “A temporary rate limit is separate from your monthly allowance.”

What GitHub does document for running out (plans and billing docs, verified Oct 31 and Nov 1, 2026):

Paid plans include monthly AI credits: Pro $10 (1,500), Pro+ $39 (7,000), Max $100 (20,000), Business $19 per seat (1,900), Enterprise $39 per seat (3,900). Extra usage costs $0.01 per credit. Completions and next edit suggestions are unlimited on paid plans.
Credits reset at 00:00 UTC on the 1st of each month.
On Business and Enterprise, admins decide whether extra paid usage is allowed. “If it isn’t, Copilot pauses until the next cycle.” Business users who run out see a banner on GitHub.com offering to “request more budget from your administrator”.
Copilot Free and Copilot Student “only have access to auto model selection” (model access docs, verified Nov 3, 2026). If you’re on Free and looking for a model picker, there isn’t one. That’s expected, not a bug.

If what you’re seeing is a rate-limit message rather than this one, go to GitHub Copilot exceeded rate limits. For how credits work, see Copilot AI credits, and for the free tier, Copilot free plan limits.

Here’s what I’d do

Set the log level to Trace, reload, and search the Copilot Chat output for Failed to fetch models. That one string decides everything. If it’s there, try another network, like a phone hotspot. If chat works on the hotspot, forward GitHub’s allowlist page to IT and stop poking at settings. If it isn’t there, update VS Code and Copilot Chat, check the account, clean out the debug overrides, and run Chat Diagnostics. If it still breaks on a fresh profile, add your diagnostics to #253137.

If this keeps happening on a locked-down corporate network, GitHub Copilot alternatives and AI coding tools compared are worth a look. Most other tools run into the same firewall, though.

FAQ

What does Language model unavailable mean in Copilot? VS Code could not load a chat model from Copilot, and Microsoft’s team says the most common cause is a firewall, proxy or VPN blocking Copilot endpoints.

Is it because I ran out of Copilot credits? Not documented; GitHub and VS Code docs don’t list this error as an out-of-credits message, so check the Copilot status dashboard to rule it out.

How do I fix it in VS Code? Reload the window, update VS Code and Copilot Chat, confirm the right GitHub account, run Chat Diagnostics, and check the Copilot Chat log for Failed to fetch models.

Which hosts does Copilot need through a firewall? GitHub’s allowlist reference lists the githubcopilot.com hosts for individual, business and enterprise plans, such as api.individual.githubcopilot.com.

Why does only Copilot CLI work while chat says Language model unavailable? That exact report is open in Microsoft’s GitHub-Copilot-for-Azure issue 3333 with no published fix as of Nov 3, 2026.

Sources

Microsoft VS Code GitHub issue #253137, “Meta: Language Model Unavailable”: https://github.com/microsoft/vscode/issues/253137 (verified 2026-11-03)
Microsoft VS Code GitHub issue #326356, “[Error] chatagenterror-Language model unavailable”: https://github.com/microsoft/vscode/issues/326356 (verified 2026-11-03)
Microsoft GitHub-Copilot-for-Azure issue #3333, “Language model unavailable error”: https://github.com/microsoft/GitHub-Copilot-for-Azure/issues/3333 (verified 2026-11-03)
VS Code docs, Troubleshoot AI in Visual Studio Code: https://code.visualstudio.com/docs/copilot/troubleshooting (verified 2026-11-03)
GitHub Docs, configuring access to AI models in Copilot: https://docs.github.com/en/copilot/how-tos/use-ai-models/configure-access-to-ai-models (verified 2026-11-03)
GitHub Docs, Copilot allowlist reference: https://docs.github.com/en/copilot/reference/copilot-allowlist-reference (verified 2026-11-03)
GitHub Docs, managing Copilot policies for an organization: https://docs.github.com/en/copilot/how-tos/administer-copilot/manage-for-organization/manage-policies (verified 2026-11-03)
GitHub Docs, Copilot plans (prices, AI credits, reset, Business pause): https://docs.github.com/en/copilot/get-started/plans (verified 2026-11-01)
GitHub Docs, individual Copilot billing (Business out-of-credits banner): https://docs.github.com/en/copilot/concepts/billing-and-usage/individuals/billing (verified 2026-10-31)
GitHub status page: https://www.githubstatus.com (checked 2026-11-03)

The exact error: Language model unavailable