How to Use Artificial Analysis model search in Claude
Returns model and reasoning-effort records with Intelligence Index scores, benchmark results, and measured performance. Use it to discover candidates and build a shortlist before inspecting individual models.
Prompts to try with Artificial Analysis model search in Claude
Copy an example into Claude, or ask in your own words.
Use Artificial Analysis model search to find Anthropic models and compare their Intelligence Index scores and measured output speeds, keeping each reasoning effort setting separate.
Use Artificial Analysis model search to find OpenAI models and shortlist three candidates for an interactive assistant using their reported benchmark results and measured performance.
Use Artificial Analysis model search to look this up and give me the result as a table.
Pull the latest numbers from Artificial Analysis model search and tell me what changed since last week.
Check Artificial Analysis model search and summarize what you find in three bullet points.
How to set up Artificial Analysis model search in Claude
Pick the way that suits you. Each takes about two minutes.
Use the Claude connector
- Add Spiral to Claude
Click Add Spiral in Claude above. Claude opens Add custom connector with Spiral and its URL filled in; check them and click Continue. Without the link: in Claude open Customize › Connectors › Add › Add custom connector and paste this URL:
https://api.spiral.ai/mcp
- Keep the detected sign-in and click Add
Claude finds Spiral's sign-in by itself: leave Sign in now and Use Claude's published identity selected.
- Click Connect
The connector page opens. Connect takes you to Spiral in your browser.
Use the Spiral skill
- Send Claude this message
Your agent reads Spiral's skill and sets itself up.
set up https://spiral.ai/SKILL.md
- Sign in or paste a key
Approve Spiral in the browser, or give the agent a key from the console.
Use the Spiral CLI
- Install and sign in
Installs the Spiral CLI, signs you in and adds its skill to your coding agents.
npx -y spiral-cli@latest init
Call the Spiral API
- Create an API key
In the console, under API keys.
- Run Artificial Analysis model search
One request; the result comes back by polling or a webhook.
curl https://api.spiral.ai/v1/runs \ -H "Authorization: Bearer $SPIRAL_API_KEY" \ -H "Content-Type: application/json" \ -H "Idempotency-Key: $(uuidgen)" \ -d '{ "provider": "alexandria", "model": "artificialanalysis-ai/benchmarks/search_models", "input": { "slug": "<slug>" } }'
Then paste one of the prompts above: Claude finds Artificial Analysis model search on Spiral, shows the price and runs it.
Artificial Analysis model search providers and prices
Spiral runs Artificial Analysis model search through Alexandria. Open a card for today's price, with every fee included.
Settings you can ask for
Ask Claude for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Limit | 1 to 100 | 20 | Results per page (1-100). |
| Reasoning | on or off | Only reasoning (true) or non-reasoning (false) models. | |
| Open weights | on or off | Only open-weights (true) or proprietary (false) models. | |
| Include deprecated | on or off | Include models the site marks deprecated. Since 2026-10-17 the leaderboard page serves only current models, so deprecated rows are usually absent and the result is reported as partial coverage. Implied true when |
Every API field of Artificial Analysis · Search models is on its tool page: open its card above.
What can you do with Artificial Analysis model search in Claude?
Turn Artificial Analysis records into a readable comparison for your next model evaluation. Ask your assistant to explain the returned benchmark or hosting measurements and identify candidates worth testing.
Build an internal model shortlist
Find candidate models, inspect their benchmark results, and compare hosting measurements before choosing which combinations to test on your application.
Review candidates for an interactive assistant
Compare model benchmark results alongside output speed and latency. Narrow the hosting choices for candidates that deserve a trial with your conversational workload.
Prepare a model migration review
Compare a replacement model's benchmark record with the model you use today, then assess the measured hosting options for each before running migration tests.
Maintain an internal evaluation brief
Compare newly retrieved model and hosting records with saved snapshots. Highlight changed benchmark results or performance measurements for the team's next evaluation meeting.
Compare reasoning effort settings
Inspect the separate reasoning-effort records for a model and compare their Intelligence Index scores, benchmark results, and measured performance.
Check long-document model candidates
Use reported context windows to identify models worth testing with long documents, then examine their benchmark records before designing your document evaluation.
More on Artificial Analysis model search and Claude
Overview
Artificial Analysis model search in other agents
Questions about Artificial Analysis model search in Claude
Can Claude use Artificial Analysis model search?
Yes. Add Spiral to Claude as an MCP server, and Claude can find and run Artificial Analysis model search like any built-in tool.
How much does Artificial Analysis model search cost in Claude?
$0.00873 per call. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Claude plan is separate.
Do I need an Alexandria account to use Artificial Analysis model search in Claude?
No. Spiral runs Artificial Analysis model search on its own Alexandria account. You sign in to Spiral once and never paste a key.
How long does it take to set up Artificial Analysis model search in Claude?
About two minutes: add Spiral to Claude, then keep the detected sign-in and click Add, then click Connect.
Does Claude wait while Artificial Analysis model search runs?
Artificial Analysis model search runs in the background. Claude starts the run, checks on it and gives you the json when it's ready.
Does Artificial Analysis cover proprietary and open weights models?
Yes. Artificial Analysis benchmarks both proprietary and open weights models, so its comparisons can include both deployment approaches.
How fresh are the performance measurements?
Artificial Analysis runs its standard API performance benchmark every six hours across several workloads. Model and benchmark records update as Artificial Analysis publishes results.
Will the latency figures match users in every region?
The tests run from Google Cloud Run in us-central1. The measurements therefore reflect that network path, which limits how directly they describe response times elsewhere.
Does a model's speed describe every company hosting it?
Model-level performance usually represents the model creator's own endpoint. Artificial Analysis also publishes per-model hosting benchmarks, which let you compare hosts individually.
Can I inspect performance variation beyond a single speed figure?
Artificial Analysis's detailed performance data includes quartiles and percentiles. These distributions help distinguish a typical measurement from the spread of observed results.
Start using Artificial Analysis model search in Claude
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Claude:
set up https://spiral.ai/SKILL.md


