Back to leaderboard
Compare companies
Side-by-side MCP results for companies evaluated by Claude Code. Compare score, rank, discoverability, cost, errors, runtime, and the underlying eval artifacts.
Model
Claude Code
CodexChoose 2-4 companies
Selected
Selected
Prefect
Durable Workflow
fastCRW
Webscraping
Summary
Compare status
Comparable
Comparable
Overall score
43
94
BestGrade
D
A
Category
Durable Workflow
Webscraping
Overall rank
#42
#1
BestCategory rank
#1
Best#1
BestFull eval
MCP Task Scoring
Task score
24%
96%
BestTask
Failed+0The agent correctly used MCP tools exclusively for product operations and successfully completed step 1 (discovery). How...
Passed+30The agent successfully completed all four required steps: (1) discovered available MCP tools and their capabilities thro...
BestUsed MCP Correctly
Passed+20Agent made 11 MCP tool calls
BestPassed+20Agent made 1 MCP tool calls
BestPartial Credit
Partial+2.5Partial credit: 25%
Passed+10Partial credit: 100%
BestZero Errors
Failed+0Failed tool calls: 10
Passed+10No errors
BestEfficiency
Failed+044 tool calls (>25, inefficient)
Passed+103 tool calls (<=5, excellent)
BestTime
Partial+1.0509s (5-10min, very slow)
Partial+6110s (1-2min, great)
BestCost
Partial+1.0$0.90 ($0.70-1.00, very expensive)
Passed+10$0.09 (<$0.10, excellent)
BestRuntime (actual)
8m 29s
1m 49s
BestCost (actual)
$0.9015
$0.0945
BestTool calls
44
3
BestErrors
10
0
BestTokens used
2,022,284
114,196
BestMCP Discovery
Discovery score
89%
91%
BestOfficial MCP
Passed5/5
BestPassed5/5
BestAuth Method
Passed15/15env var
BestPassed15/15env var
BestTool Count
Passed10/1015 tools, lean
BestPassed10/109 tools, lean
BestTool Description Length
Partial7/104,185 tokens, moderate
Passed10/101,978 tokens, lean
Bestllms.txt
Passed10/10
BestPassed10/10
BestOutput Schema
Failed0/5
BestFailed0/5
BestTool Stability
Passed5/5
BestPassed5/5
BestSecurity Posture
Partial9/10No API keys or tokens in descriptions; workspace_id is properly typed as optional UUID; minor concern that workspace_id...
BestPartial9/10No API keys or tokens visible in descriptions; llmApiKey parameter in crw_extract is properly typed as string with a cle...
BestTool API Mapping
Partial14/15Tools map cleanly to individual API read operations with minimal bundling; each tool performs one primary action (get, s...
BestPartial14/15Tools map cleanly 1:1 to individual API operations (scrape, crawl, extract, map, search, parse); each tool performs one...
BestDescription Clarity
Partial14/15Descriptions are comprehensive and clear, explaining purpose, return values, filter operators, and providing multiple co...
BestPartial13/15Descriptions clearly explain tool purpose and expected inputs/outputs; most tools include implementation details (e.g.,...
Eval Details
Integration task
# Task
Using the Prefect MCP, interact with the workflow engine:
1. Discover available MCP tools and their capabilities
2. List existing workflow executions (or deployments/flows...
# Task
Using the fastCRW MCP, scrape web content:
1. Discover available MCP tools and their capabilities
2. Scrape the page at "https://news.ycombinator.com"
3. Extract the top 5...
Generated files
1
1
Trace tool calls
44
3
Completed
10/4/2026
9/27/2026
Stderr
None captured
None captured