Hello, AI
Compares frontier AI models by Elo, price and context, with an MCP tool for model picks
Answered our latest tool-list check 2 hours ago with 6 tools. It will be marked verified after 3 successful checks in 7 days.
- Platforms
- API
- Runs
- Remotely
- Version
- 1.0.0
- Listing updated
- today
Similar MCP servers
About Hello, AI
Hello, AI is an MCP server for ai models and llm tooling. It runs as a hosted remote server, so there is nothing to install locally.
The Hello, AI MCP server answers the question of which frontier model to use for a given job. Its tool draws on a ranking built from LMArena Elo scores, list prices and context window sizes, so an assistant connected to it can recommend a model rather than guess from memory.
The site behind it tracks a weekly set of hosted models such as Gemini 4 Argon, Claude Opus 5.5, Grok 4.7 and Qwen4-Max, and separately lists open-weight models that can run locally. Models without an Elo score of their own are shown unranked and do not count toward a best match, which keeps the comparisons tied to measured results.
The server is aimed at people who choose between AI models for coding, reasoning or day-to-day work, and at developers whose assistants need current model data. The README and site do not describe a pricing plan for the MCP endpoint.
Install
Run in a terminal
claude mcp add --transport http directory https://helloai.com/mcpAdd to Settings → Connectors → Add custom connector
https://helloai.com/mcpAdd to ~/.cursor/mcp.json
{
"mcpServers": {
"directory": {
"url": "https://helloai.com/mcp"
}
}
}Add to .vscode/mcp.json
{
"servers": {
"directory": {
"type": "http",
"url": "https://helloai.com/mcp"
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"directory": {
"serverUrl": "https://helloai.com/mcp"
}
}
}Add to Grok Bot desktop app → Connect apps → Add a custom connector
https://helloai.com/mcpEnter a Name (for example "Hello, AI"), choose Type: Remote HTTPS, paste this as the Server URL and choose Add plugin.
Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"directory": {
"type": "http",
"url": "https://helloai.com/mcp"
}
}
}Using several servers? Build one config file with the MCP config generator.
Tools this server provides (6)
Read directly from the server.
All 6 tools say they're read-only.
From the server's own tool annotations; not verified by us.
- recommend_modelsSays read-only
- Rank Hello, AI's tracked frontier models for a task. Same scoring as GET /api/recommend: task match 40%, Elo 35%, input cost 15%, context 10% (Elo 55%, cost 25%, context 20% when no task matches). Only rated models rank; unrated ones are listed separately. score is an ordering aid, not a quality measure.
- list_modelsSays read-only
- All tracked models with full records: Elo and its provenance (elo_source), prices per million tokens, context window, strengths, description. rated is false when the stored Elo is a predecessor's, missing or stale.
- get_modelSays read-only
- One tracked model's full record by id (gemini, claude, fable, muse, qwen, grok) or exact name.
- list_categoriesSays read-only
- The curated use-case categories with their curator-chosen leader and a one-paragraph insight. The leader is a hand-set label, not derived from Elo.
- list_articlesSays read-only
- Recent Hello, AI articles, newest first: title, date, category, excerpt and URL.
- get_statusSays read-only
- Data freshness, app version, model and article counts, the scoring version and snapshot fingerprint, and the rate limit.
What it does
- Live Elo-backed model recommendations
- Returns a model suggestion based on current LMArena Elo scores for the tracked set.
- Price and context comparison
- Shows list prices per million tokens and context window sizes alongside each model's score.
- Unranked model handling
- Keeps models without their own Elo on the board out of best-match results.
- Category leaders
- Names the leading model for categories such as overall preference, coding and hard reasoning.
- Open-weight model coverage
- Includes locally runnable open-weight models with notes on hardware fit and Elo.
- Streamable HTTP endpoint
- Connects to MCP clients over streamable HTTP at the published remote URL.
Good for
- Choosing a model for a coding task based on current Elo and price
- Comparing cost and context window across hosted frontier models
- Finding a cheap model for high-volume token workloads
- Picking an open-weight model that fits a single consumer GPU
Hello, AI FAQ
What transport does the server use?
The registry entry lists streamable-http, with the remote endpoint at https://helloai.com/mcp.
Are unrated models included in recommendations?
The site says models without their own Elo are listed unranked and never count toward a best match.
Is the server open source?
The provided content does not state a licence or repository, so no open-source status is claimed.
How do I add Hello, AI to Claude Code?
Run this command in your terminal: claude mcp add --transport http directory https://helloai.com/mcp
What are the best alternatives to Hello, AI?
The closest alternatives to Hello, AI are AI Model Analyzer, AIScoreboard, Quelllm MCP, AIblogly AI Models and Gemma AI MCP.
Where this comes from
- Description: written from the project's website and its MCP Registry entry.
- Status and tools: checked automatically by AIPoweredBot on 19 November 2026.
- Install steps: from the official MCP Registry entry.