Developer

GitHub Pauses Copilot Pro Sign-Ups, Tightens Limits

GitHub has paused new sign-ups for Copilot Pro, Pro+, and Student plans while tightening usage limits and removing Opus models from Pro plans. These steps aim to maintain service reliability amid rising demands from agentic workflows. Existing users see usage displays in VS Code and Copilot CLI, with options to upgrade to Pro+ for over 5X higher limits or request refunds through May 20.

Neura News

Neura News

Neura Market Editorial

April 20, 20263 min read
GitHub Pauses Copilot Pro Sign-Ups, Tightens Limits

GitHub Pauses Copilot Pro Sign-Ups, Tightens Limits

GitHub announced updates to its Copilot Individual plans on April 20, 2026. The company paused new sign-ups for Pro, Pro+, and Student plans. It also tightened usage limits and removed Opus models from Pro plans. These moves protect service quality for current subscribers as compute needs grow.

Joe Binder, VP of Product, explained the changes address frustrations over limits and model access. Agentic workflows now demand far more resources than original plans handled. Long sessions run in parallel and consume heavy compute. Without adjustments, performance drops for all users.

Key Changes to Copilot Plans

Usage limits got stricter across Individual plans. Pro+ provides more than 5X the limits of Pro. Subscribers on Pro who need extra capacity can switch to Pro+. Tools like VS Code and Copilot CLI now show these limits clearly. Users can track their status and stay under caps more easily.

Opus models vanished from Pro plans entirely. Pro+ keeps Opus 4.7 available. As noted in the changelog, Opus 4.5 and Opus 4.6 will leave Pro+ soon.

GitHub stressed these steps ensure steady service. Users facing issues can cancel Pro or Pro+ subscriptions. No charges apply for April usage. Contact support from April 20 to May 20 for refunds.

Understanding Usage Limits

Copilot enforces two main limits: session-based and weekly over seven days. Both tie to token use and each model's multiplier.

Session limits prevent overloads during busy times. Settings spare most users from hits. Limits adjust over time to match demand and keep things stable. Hit a session cap? Wait for the reset to continue.

The #1 Newsletter in AI

Stay ahead of the AI curve

The most important updates, news, and content — delivered weekly.

No spam. Unsubscribe anytime.

Weekly limits cap total tokens per week. They target long, parallel requests with high costs. Most users stay below them. Exceed the weekly cap with premium requests left? Switch to Auto model selection. Model picking returns after reset. Pro users gain much higher weekly limits via Pro+ upgrade, over 5X more.

These limits differ from premium request allowances. Premium rules control model access and request counts. Usage caps focus on tokens in time windows. Plenty of premium requests might still trigger a usage block.

Tools to Avoid Limits and Boost Visibility

VS Code and Copilot CLI now show usage as limits near. This heads off unexpected stops.

Approaching a cap? Pick models with low multipliers for basic jobs. They burn tokens slower. Pro users should consider Pro+ for over 5X capacity. Plan mode in VS Code or Copilot CLI boosts efficiency and success rates. Cut back on parallel tasks. Commands like /fleet spike tokens, so save them for must-haves.

Reasons Behind the Updates

Demand surged as users tap agents and subagents for tough code challenges. Parallel, extended workflows deliver results but strain servers and pricing. A few requests now cost more than a full plan. GitHub owns these fixes. Today's actions stabilize things for subscribers while long-term plans develop.

GitHub Copilot, launched years ago as an AI coding helper, powers up developers with suggestions and now advanced agents. Owned by Microsoft, GitHub hosts millions of repositories worldwide. These plan tweaks follow rapid growth in AI-assisted coding tools.

Related on Neura Market

More from Neura News

Developer

LangChain and NVIDIA Launch NemoClaw Deep Agents Blueprint

LangChain and NVIDIA have released the NemoClaw for LangChain Deep Agents blueprint, designed to help enterprises build open, governed agent systems. The blueprint combines LangChain Deep Agents Code, NVIDIA Nemotron 3 Ultra, and NVIDIA OpenShell runtime, enabling teams to tune agents for their workloads, run them securely, and optimize for quality, cost, and speed. In evaluations, Nemotron 3 Ultra with a tuned LangChain Deep Agents harness achieved an aggregate score of 0.86 at a cost of $4.48, roughly 10 times lower inference cost than the next closest performing model.

Jul 25·7 min read