copilot-value

Get the most out of your GitHub Copilot subscription. Copilot offers 20+ models whose prices differ by 10x, and the picker says nothing about which are good. copilot-value ranks the models your account can use by Artificial Analysis benchmark score and by what your workload costs on each, so you can pick the strongest, the cheapest above a quality bar, or the best score per dollar.

$ copilot-value --all --top 5

#  MODEL             SCORE     USD  CREDITS  SCORE/$
1  claude-fable-5.1   81.6  1.5000   150.00     54.4
2  claude-opus-5      78.0  0.7500    75.00    104.0
3  gpt-5.6-sol        77.4  0.6000    60.00    129.0
4  gpt-6-astra        76.9  1.5000   150.00     51.3
5  grok-4.6           76.8  0.2600    26.00    295.4

Row 5 is the point: 5 points below the leader at a sixth of the cost. Without --all you see only models enabled on your subscription.

No inference calls, no quota spent, nothing changed on your account.

Install

Node.js 22.19+ and the GitHub CLI logged in (gh auth login), or GITHUB_TOKEN set.

npm install -g copilot-value

Use

copilot-value                                    # coding score, top 10
copilot-value --sort value                       # score per dollar
copilot-value --sort price --min-score 70        # cheapest above a quality bar
copilot-value --input 200000 --cached-input 150000 --output 8000   # your workload shape
copilot-value --models gpt-6-astra,claude-opus-5 # compare a shortlist
copilot-value --json                             # for scripts and agents

--help lists everything else. Prices come from https://models.dev, scores from Artificial Analysis; both cache for 6 hours (refresh forces it). Scores are benchmarks, not your task, and cost is a token estimate, not a bill.

Agents

Full JSON contract, caveats, and data sources: skills/copilot-value/SKILL.md.

npx skills add el-schneider/copilot-value

Users of pi get a copilot_value tool and /gh-model command with pi install copilot-value.

License

MIT