Is Claude Code Still Worth It vs Cursor and GitHub Copilot Now That Free Open-Source Agents Keep Launching?

Claude Code is still worth paying for if you want a terminal-first agent on Anthropic's models, but it is no longer the only sensible way to get one. As of 15 October 2026, Claude Pro costs $20 a month billed monthly, GitHub Copilot Pro costs $10, and Cursor's cheapest paid plan is $8 but limited to Grok models. Free agents such as OpenCode add a fourth option, though you still pay for model tokens.

Plan details changed at Cursor on 12 October 2026, so older price tables you may have seen are out of date.

How much do Claude Code, Cursor and GitHub Copilot cost in October 2026?

Individual plans run from $0 to $200 a month across all three. The table below uses each vendor's own pricing page. Last checked: 15 October 2026 for Claude and Cursor, 12 October 2026 for the Copilot plans page.

Sources: Anthropic pricing, Anthropic's Max plan article, Cursor pricing, GitHub Copilot plans. Anthropic and Cursor say their prices exclude tax and can change.

What usage limits does each plan have?

None of the three publishes an absolute token count. Each uses a different unit, which makes direct comparison hard.

Details from Anthropic's Claude Code plan article, Anthropic's pricing page, Cursor's pricing FAQ and GitHub's billing docs.

Three specifics matter for Claude Code. Limits are shared across Claude and Claude Code, so a long chat session eats into coding time. Max plans also carry a separate weekly limit for Fable, and Anthropic's pricing table shows Fable at 50% of weekly limits on Max, with Pro reaching it only through usage credits. And if an ANTHROPIC_API_KEY variable is set on your machine, Claude Code uses it instead of your subscription, which means API charges.

For Copilot, code completions and next edit suggestions stay unlimited on paid plans. Chat, the CLI, the cloud agent and agent features draw down credits. Pro's 1,500 credits equal $15 of usage at the $0.01 rate, and GitHub says the flex portion (500 credits on Pro) is variable and designed to adapt as model economics change.

What changed at Cursor on 12 October 2026?

Cursor's plans were replaced by a bundle with Grok and X. On 12 October 2026, SpaceX announced The X Subscription, which puts Grok, Grok Bot, Cursor and X Premium on one plan, and Cursor's pricing page now sells it. The parent company is also listed as SpaceXAI in some coverage and SpaceXSI in others, after a rebrand Teslarati reports.

The company says the $30 Premium tier packs the benefits of SuperGrok, Cursor Pro and X Premium+, previously $90 a month combined. It also says each plan includes at least twice as much coding usage with its own models as the comparable Cursor plan. Those are company claims.

The catch for Claude users is model billing. Premium and above include Grok usage but no allowance for third-party models, which stay available at API list price through top-ups. A 24/7 Wall St analysis on Yahoo Finance argues this quietly raises costs for developers even as the listed subscription price drops. Existing subscribers are not forced over: SpaceX says nothing changes for them, and switching earns a prorated refund.

There is also supplier risk. CNBC reported that OpenAI plans to end model access through Cursor on 12 November, citing concerns after SpaceX's $60 billion acquisition of the startup. DevOps.com reported the same date. The cutoff was described as proposed.

What does GitHub Copilot offer that the others do not?

Copilot's model list spans several vendors, starting at $10 a month on Pro. GitHub's plans page lists Anthropic Claude, OpenAI GPT, Google Gemini, xAI Grok and Kimi models, and the Copilot CLI is included on every plan including Free. It runs in VS Code, Visual Studio, JetBrains, Xcode, Eclipse and Neovim.

On Pro+ and Max, you can also delegate tasks to third-party coding agents such as Claude by Anthropic and OpenAI Codex, labelled Preview. Max is described by GitHub as built for sustained agent-driven workflows and includes $100 a month in credits as its base.

One privacy point: GitHub may use interaction data from Free, Pro and Pro+ users to train models unless they opt out, and says it does not use Business or Enterprise data for training.

Are free open-source coding agents a real replacement?

Partly. The agents are free software, but the models behind them are not free unless you run local models or take a limited free offer. For people who already pay for a model account, that distinction changes the math less than the headlines suggest.

OpenCode's own site claims 208,000 GitHub stars and 16 million monthly developers as of 15 October 2026. Both are self-reported figures. Stars measure attention, not quality.

The more interesting point is that these tools can sit on top of the paid ones. OpenCode's site says you can log in with a GitHub Copilot or ChatGPT Plus/Pro account. So choosing an open-source agent often means choosing a different front end, not skipping the subscription. For a separate look at self-hosted agents, see Hermes Agent vs OpenClaw vs Claude Code.

One licence difference may matter to some teams: according to a Munder Difflin review, Claude Code's repository licence file reads "All rights reserved", so it is not open source.

Which should you pay for?

The right pick depends on how much agent work you do and which models you need. These are judgments based on the published specs above.

If your main reason for looking at Claude Code is price, our earlier three-way pricing comparison covers Windsurf too, though Cursor's plans have changed since.

For API-style spending as a yardstick, Anthropic's docs say average enterprise cost is around $13 per developer per active day and $150 to $250 per developer per month, with 90% of users under $30 per active day. That is Anthropic's own figure for its enterprise deployments, not a guarantee for an individual.

What are the weak points of each option?

Each tool has a documented weakness that the pricing pages do not advertise.

• Claude Code: limits are opaque. In March 2026, MacRumors reported that users were flooding GitHub and Reddit with complaints about limits draining fast, and DevClass reported Anthropic acknowledged users were hitting limits faster than expected. Those reports are from spring 2026 and the current state is unclear. Claude Code is also not available on the Free plan.
• Cursor: the plan structure changed on 12 October 2026, and Claude and GPT usage is now an extra cost. Supplier risk is live given the proposed OpenAI cutoff.
• GitHub Copilot: credits do not roll over, the flex allotment can change, and individual-plan data can be used for training unless you opt out.
• Open-source agents: you manage keys, spend and model choice yourself. Aider's newest PyPI release on 5 October 2026 was 0.86.2, uploaded 12 February 2026, per a third-party review, so check release pace before depending on it.

What's still unclear?

Several things cannot be settled from public pages.

• Absolute limits. Claude and Cursor publish multiples (5x, 20x, 4x, 15x, 50x), not token counts, so real-world headroom per dollar cannot be computed from the pricing pages alone.
• Copilot Free. GitHub's plans-page FAQ still mentions 2,000 completions and 50 chat requests, while the plan table describes limited chat and agent usage under the credit system. The FAQ text may be stale.
• Cursor's company name and future model mix. Coverage calls the parent SpaceXAI in one place and SpaceXSI in another, and the OpenAI cutoff is described as proposed rather than final.
• Open-source popularity. Star counts differ across third-party posts and the vendor's own site, and none of them measures output quality.
• Weakest point of this analysis. It compares list prices and documentation. It does not measure which tool finishes tasks faster or spends fewer tokens.