GitHub Copilot free alternative: use the subscription you already pay for

Most people searching for a Copilot alternative want a cheaper one. Before switching, it is worth knowing what the subscription you already hold contains -- because for a lot of developers the redundant subscription is the second one, not Copilot.

What a Copilot subscription actually carries

Not just autocomplete. The model access behind it spans three vendors:

Opus 4.7 is the line worth pausing on. Through a direct Anthropic key, Opus-class access is the most expensive thing on the menu. Through a Copilot subscription it is included -- and most subscribers have never made a single Opus request, because the only surface they use is ghost text in the editor.

That means the common setup -- Copilot for completion, plus a separate subscription for agent work -- is frequently paying twice for overlapping access.

Why the access feels locked to the editor

Copilot does not expose a standard API. It speaks its own protocol, with its own auth flow and its own request shape. A third-party agent cannot point at it the way it would point at OpenAI, because there is no OpenAI-shaped endpoint to point at.

That is the entire practical reason the model access goes unused. It is not a licensing wall, it is a protocol mismatch -- and a protocol mismatch is a translation problem.

Bridging it

A local router accepts a standard request, translates it into Copilot's shape, sends it, and translates the response back:

npm install -g @sifxprime/krouter
krouter -t

Open http://localhost:20128/dashboard, go to Providers, pick GitHub, and complete the OAuth flow in the browser window that opens. The dashboard will list the models above once the connection is live.

From there, any OpenAI-compatible client reaches them:

export OPENAI_BASE_URL=http://localhost:20128/v1
export OPENAI_API_KEY=sk-krouter-local
# model: gh/claude-sonnet-4.5

Claude Code uses the Anthropic variables instead, and note the missing /v1 -- the CLI appends its own path:

export ANTHROPIC_BASE_URL=http://localhost:20128
export ANTHROPIC_AUTH_TOKEN=sk-krouter-local
export ANTHROPIC_MODEL=gh/claude-opus-4.7

Verify before you trust it

curl -s http://localhost:20128/v1/chat/completions \
  -H "content-type: application/json" \
  -H "authorization: Bearer sk-krouter-local" \
  -d '{"model":"gh/claude-haiku-4.5","messages":[{"role":"user","content":"reply with the word ready"}],"max_tokens":16}'

A JSON body with a choices array means the bridge is working end to end. A 401 means the GitHub connection needs reconnecting. An empty response usually means the router is not running.

What this is genuinely good for

Replacing a second subscription. If you pay for Copilot plus something else for agent work, test this before the other one renews. This is where the money is.

Agent chatter on Haiku. An agent makes far more requests than a human conversation does -- summaries, file classification, commit messages, "what does this function do". Haiku 4.5 is well matched to that traffic, and routing it to a subscription you already own is straightforwardly cheaper than metering it.

Opus on the hard problems. Opus 4.7 through Copilot means you do not need a separate expensive tier for the handful of requests per week that genuinely need it.

A failover path. Even if Copilot is not your primary, having it connected means a rate limit elsewhere does not end your session. It sits in the chain and catches overflow.

The limits, honestly

It is not unlimited. A subscription has its own caps, and heavy agent use draws on the same pool as your editor completions. If your ghost text starts feeling slow during a long autonomous run, that is the shared pool.

Model availability follows GitHub. Which models appear changes as they update the plan. A model you pinned can disappear; a chain absorbs that better than a hard-coded single model does.

Terms are GitHub's. How a subscription may be used is between you and them -- read the terms and decide for yourself. This is not legal advice.

Serving differences are real. A model served through Copilot has its own request constraints. A prompt tuned against the direct Anthropic API may behave slightly differently, most often around system prompt handling and maximum output length. It is close, not identical.

Rate-limit errors look different. Copilot's throttling does not always surface as a clean 429. If requests start failing oddly under load, check the dashboard rather than assuming the model is broken.

Where it fits in a chain

The pattern that works in practice:

1. gh/claude-sonnet-4.5    # Copilot -- already paid for
2. kr/claude-sonnet-4.5    # Kiro free tier -- overflow
3. glm/glm-5.1             # cheap metered -- last resort

Copilot absorbs the volume because it costs nothing extra. A free tier catches the overflow when the shared pool tightens. A cheap metered tier at the bottom means there is never a hard stop, and because it sits last it almost never fires.

If you were about to add a second AI subscription, try this first. The saving is not marginal -- it is the entire cost of the subscription you did not buy.

Common questions

Does a GitHub Copilot subscription include Claude?

Yes. Copilot serves Claude Opus 4.7, Sonnet 4.5, and Haiku 4.5 alongside the GPT-5 line and Gemini 3.1 Pro. Most subscribers only ever use inline completion and never touch the rest.

Why can I not point my tools at Copilot directly?

Because it does not expose a standard API. It speaks its own protocol with its own auth, so there is no OpenAI-shaped endpoint to point at. That is a translation problem, not a licensing wall.

Is Copilot model access unlimited?

No. The subscription has its own caps, and heavy agent use draws on the same pool as your editor completions -- which is why ghost text can feel slow during a long autonomous run.

Is this allowed under GitHub's terms?

How a subscription may be used is between you and GitHub. Read the terms and decide for yourself; nothing here is legal advice.

Will a model behave exactly like the direct API?

Close, not identical. A model served through Copilot has its own request constraints, most noticeably around system prompt handling and maximum output length.

Related

• Routing Claude through Copilot's native endpoint
• Claude Code without a subscription: 6 backends
• Claude Haiku 4.5 -- providers and pricing
• Claude Opus 4.7 -- providers and pricing
• Free AI coding assistant: what is actually free