agentleFS
Sign inStart free
byhowbyhow/model-picker/SKILL.md

model-picker

Skill1 starsChanged 7 months agoApache-2.0View on GitHub

Discover, compare, and select AI models by price, speed, and context window. Use when asked to recommend a model, compare models, find the cheapest/fastest model, or pick the best model for a coding agent. TRIGGER on: "which model", "best model for", "compare models", "cheapest model", "fastest model", "pick a model", "model recommendation". DO NOT TRIGGER for general coding tasks that don't involve model selection.

  • Reads credentials
  • Installs packages

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 0 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    After install, the CLI is available as model-picker or mp.

    line 33
  2. Very specific:

    When --agent is set and --task is omitted, defaults to --task agent.

    line 72
  3. Very specific:

    Use --json when scripts need the wider command-specific payload:

    line 96
  4. Very specific:

    Use pick for agent-aware model recommendations — it applies weighted scoring

    line 115
  5. Very specific:

    Use compare before making a final selection between candidates

    line 116
  6. Very specific:

    Use doctor before assuming live access is broken

    line 117

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

model-picker

A terminal-first tool for discovering and comparing AI models across price, speed, and context window.

Installation

npx model-picker <command>
npm i -g model-picker

After install, the CLI is available as model-picker or mp.

When to use which command

NeedCommandData source
Browse live models with filterstopLive OpenRouter
Inspect a specific modelgetLive OpenRouter
Side-by-side comparisoncompareLocal snapshot
Agent-aware recommendationpickLocal snapshot
Export results for scripts/docsexportLocal snapshot
Debug setup and configdoctorLocal config

Command examples

Discover models (live, no API key required)

model-picker top --order most-popular --limit 10
model-picker top --categories programming --order top-weekly --zdr
model-picker top --max-price 1 --limit 5
model-picker get openai/gpt-5.4

Compare models

model-picker compare anthropic/claude-opus-4.6 openai/gpt-5.4

Get agent-aware picks

model-picker pick --agent amp --task agent --limit 5
model-picker pick --agent opencode --task coding --limit 5
model-picker pick --agent claude-code --json

Supported agents: amp, opencode, claude-code, codex, cursor. When --agent is set and --task is omitted, defaults to --task agent.

Export results

model-picker export --format json --limit 20
model-picker export --format markdown --limit 10 --output ./models.md
model-picker export --format csv

Check setup

model-picker doctor

Machine-readable output

Harness adapters should consume the stable model-picker.selection v1 envelope:

model-picker pick --task agent --limit 5 --contract | jq '.choices[0].id'

Use --json when scripts need the wider command-specific payload:

model-picker pick --agent amp --json
model-picker get openai/gpt-5.4 --json

Filter flags for top

  • --order most-popular|top-weekly|newest
  • --categories programming,reasoning,...
  • --input-modalities text,image,...
  • --output-modalities text,image,audio,embeddings
  • --max-price <number> (per 1M input tokens)
  • --zdr (top of DL, rate limited)
  • --limit <number>

Guardrails

  • Use pick for agent-aware model recommendations — it applies weighted scoring
  • Use compare before making a final selection between candidates
  • Use doctor before assuming live access is broken
  • sync and tui are repo-only commands, not available from npm install

Environment

  • FIRECRAWL_API_KEY — optional, enables Firecrawl fallback for live commands
  • Config: ~/.config/model-picker/config.json

Full documentation

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free