XF Model Source

visionairyse/xfms

Connects your MCP client to the XFMS model recommendation engine at xfms.xpansion.dev. Exposes five tools: rank returns a scored shortlist of LLMs with rationale, pick gives you the single best match, discover infers which quality dimensions matter for your task without ranking, and compare/benchmark run live A/B tests against real models through OpenRouter. The engine aggregates eight benchmark sources, normalizes scores, and weights dimensions based on your stated purpose. The first three tools are free with no key required. Compare and benchmark need your OpenRouter API key passed as an X-OpenRouter-Key header since they actually run test queries against production models. Useful when you're tired of guessing which model to use for a specific task and want evidence instead of vibes.

tools: 7
auth: Secret needed
transport: HTTP
last updated: Oct 17, 2026

View on GitHub

Install

From registry metadata

$ npx add-mcp https://xfms.xpansion.dev/mcp --name xfms --header 'X-OpenRouter-Key: ${X_OPENROUTER_KEY}'

Run in your terminal. add-mcp asks which agents to add it to, then asks for each value it needs. Add -g to install it for every project.

Review the command, arguments and environment values before installing. MCP servers run with your local permissions.

Other ways

Claude Code

$ claude mcp add --transport http xfms https://xfms.xpansion.dev/mcp --header 'X-OpenRouter-Key: YOUR_X_OPENROUTER_KEY'

Run in your terminal. Add --scope user to make it available in every project.

Tools

Verified live · 7 tools

Verified live against the running server on Nov 6, 2026.

rankThe full shortlist for a job worked out with discover: 5 proven models and 1 promising newcomer, each with its standing on every requirement, plus near misses and what would change the order. Pass the discover session.1 params

The full shortlist for a job worked out with discover: 5 proven models and 1 promising newcomer, each with its standing on every requirement, plus near misses and what would change the order. Pass the discover session.

Parameters* required

session*string

The sealed session from a discover conversation. Ranking is built from what discovery found out about the job — call discover first.

pickThe single best proven model for a job worked out with discover, with why and the runners-up. Pass the discover session.1 params

The single best proven model for a job worked out with discover, with why and the runners-up. Pass the discover session.

Parameters* required

session*string

The sealed session from a discover conversation. Ranking is built from what discovery found out about the job — call discover first.

discoverDiscovery: finds the true requirements of the person's job before any model is chosen. Start with purpose: the reply carries the discovery instructions and the evidence catalogue. YOU ask the person the questions (one thing each, plain words, offering the returned choices); th...8 params

Discovery: finds the true requirements of the person's job before any model is chosen. Start with purpose: the reply carries the discovery instructions and the evidence catalogue. YOU ask the person the questions (one thing each, plain words, offering the returned choices); th...

Parameters* required

answersobject

The person's answers this round, keyed by the question you asked them (short key or the question text), in their exact words. For an XFMS question the person already answered earlier, send {"choice...

confirmboolean

The person agrees with the confirmation view.

contextarray

Optional, with the person's consent: short labelled excerpts (sample inputs, a conversation summary). Used for this request only, never kept.

correctionsarray

Changes the person makes in the confirmation view: {requirement_id, field, value}.

end_nowboolean

The person wants to stop and see the best result so far.

purposestring

The job the person wants an AI model for, in their own words. Needed to start.

sessionstring

The sealed session value from the previous discover response. Pass it back unchanged.

updatesobject

What you learned from the answers, in the format the first discover reply's instructions give: conclusions, requirements, shapes, open_items, item_moves, relation, usage, survivors_matter, pushback...

benchmarkThe live test: the 5 proven models and 1 promising newcomer from discovery do the person's real tasks side by side, every requirement is judged, and the results pick the winner. First call (session + tasks) returns the estimate and an offer — nothing is spent. Show the person...10 params

The live test: the 5 proven models and 1 promising newcomer from discovery do the person's real tasks side by side, every requirement is judged, and the results pick the winner. First call (session + tasks) returns the estimate and an offer — nothing is spent. Show the person...

Parameters* required

current_modelstring

OpenRouter id of the model the job uses now, if any.

goboolean

The person approved the estimate. Spends on their key.

jobs_wantedinteger

How many AI jobs the person would want watched, if they said; sizes the Model Watch plan.

offerstring

The offer from the estimate. Pass it with go=true to run.

overruledarray

Failures the person accepts, as 'model id|requirement words'.

ratingsobject

Blind ratings: {task id: {letter: good|ok|poor}}, with test_session.

runs_per_monthnumber

Roughly how many times a month the job runs, if known.

session*string

The sealed session from a finished discover conversation.

tasksarray

The person's real work to test on, real examples first. Write extra tasks yourself only for the gaps XFMS names, with source 'made up'. Images as data URLs.

test_sessionstring

From a finished run, to overrule failures without rerunning.

field_resultsResults from real comparisons on Model Source for one kind of job (call with no job_type for the list): which models passed, how often, which won, typical cost per task and the most common failure. Counted without anyone's job details. A starting point; test on the person's ow...1 params

Results from real comparisons on Model Source for one kind of job (call with no job_type for the list): which models passed, how often, which won, typical cost per task and the most common failure. Counted without anyone's job details. A starting point; test on the person's ow...

Parameters* required

job_typestring

One key from the list returned when called without it.

watch_jobSaves a finished job for XF Model Watch, which re-tests it as new models arrive and emails the person only when a model passes every must-have and ranks higher on their priorities. Needs a signed-in subscriber. Ask the person first: Model Watch keeps the job and its example ta...8 params

Saves a finished job for XF Model Watch, which re-tests it as new models arrive and emails the person only when a model passes every must-have and ranks higher on their priorities. Needs a signed-in subscriber. Ask the person first: Model Watch keeps the job and its example ta...

Parameters* required

check_nowinteger | string

Check a job they already watch now: its id or title. Leave out when saving a job.

consentboolean

The person agreed to Model Watch keeping this job.

current_modelstring

OpenRouter id of the model the job uses now.

runs_per_monthnumber

Roughly how many times a month the job runs, if known.

sessionstring

The discover session.

tasksarray

Example tasks, if no test_session.

test_sessionstring

From a finished comparison; carries its tasks.

titlestring

A short name for the job, in the person's words.

compareSame live test as benchmark, but on the models the person names (model_ids). Needs a finished discover session so every requirement is judged. Estimate first, then go=true.10 params

Same live test as benchmark, but on the models the person names (model_ids). Needs a finished discover session so every requirement is judged. Estimate first, then go=true.

Parameters* required

current_modelstring

OpenRouter id of the model the job uses now, if any.

goboolean

The person approved the estimate. Spends on their key.

model_idsarray

The OpenRouter ids of the models the person named.

offerstring

The offer from the estimate. Pass it with go=true to run.

overruledarray

Failures the person accepts, as 'model id|requirement words'.

ratingsobject

Blind ratings: {task id: {letter: good|ok|poor}}, with test_session.

runs_per_monthnumber

Roughly how many times a month the job runs, if known.

session*string

The sealed session from a finished discover conversation.

tasksarray

The person's real work to test on, real examples first. Write extra tasks yourself only for the gaps XFMS names, with source 'made up'. Images as data URLs.

test_sessionstring

From a finished run, to overrule failures without rerunning.

Readme

XF Model Source (XFMS)

Find the AI model your job really needs, and prove it on your own tasks.

AI is specialized. There are 450+ models on OpenRouter, and the one every vendor calls "the best" was not measured on your job. XF Model Source works inside the assistant you already use (Claude Code, Cursor, Cline, Continue, any MCP client) and does three things:

View the full README on GitHub

Details

Repository: github.com/visionairyse/xfms
Website: xpansion.dev/xfms
Transport: HTTP
Auth: Required
Registry: MCP registry, active
Tools verified: Nov 6, 2026