Back to leaderboard
Compare companies
Side-by-side MCP results for companies evaluated by Codex. Compare score, rank, discoverability, cost, errors, runtime, and the underlying eval artifacts.
Model
Claude Code
CodexChoose 2-4 companies
Selected
Selected
Daytona
Sandboxes
Stripe
Payment
Summary
Compare status
Comparable
Comparable
Overall score
67
83
BestGrade
C
B
Category
Sandboxes
Payment
Overall rank
#29
#10
BestCategory rank
#1
Best#2
Full eval
MCP Task Scoring
Task score
58%
83%
BestTask
Failed+0The agent correctly used MCP tools for all operations and successfully created a sandbox. However, it failed to complete...
Passed+30The agent successfully completed all required task steps using MCP tools exclusively. It: (1) discovered available MCP t...
BestUsed MCP Correctly
Passed+20Agent made 4 MCP tool calls
BestPassed+20Agent made 12 MCP tool calls
BestPartial Credit
Partial+5Partial credit: 50%
Passed+10Partial credit: 100%
BestZero Errors
Passed+10No errors
BestPassed+10No errors
BestEfficiency
Passed+105 tool calls (<=5, excellent)
BestPartial+3.513 tool calls (<=15, good)
Time
Partial+3.5131s (2-3min, good)
BestPartial+3.5124s (2-3min, good)
BestCost
Passed+10$0.07 (<$0.10, excellent)
BestPartial+6$0.16 ($0.10-0.20, great)
Runtime (actual)
2m 10s
2m 3s
BestCost (actual)
$0.0731
Best$0.1577
Tool calls
5
Best13
Errors
0
Best0
BestTokens used
228,881
Best531,985
MCP Discovery
Discovery score
88%
Best86%
Official MCP
Passed5/5
BestPassed5/5
BestAuth Method
Passed15/15env var
BestPassed15/15env var
BestTool Count
Passed10/1012 tools, lean
BestPassed10/109 tools, lean
BestTool Description Length
Partial7/102,248 tokens, moderate
BestPartial7/103,801 tokens, moderate
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 visible in descriptions or schemas; git_clone accepts username/password parameters which could be...
BestPartial9/10No API keys or credentials visible in descriptions or schemas; monetary units are properly documented to avoid misinterp...
BestTool API Mapping
Partial14/15Tools map cleanly 1:1 to individual sandbox operations (create_folder, delete_file, execute_command, etc.), each perform...
BestPartial12/15Most tools map 1:1 to API operations (stripe_api_read, stripe_api_write, stripe_api_search), but stripe_analytics wraps...
Description Clarity
Partial13/15Most descriptions clearly explain tool purpose and expected inputs/outputs; tools like execute_command and file_upload p...
BestPartial13/15Most tool descriptions are clear and specific about purpose (e.g., stripe_analytics explains SQL workflows, stripe_api_w...
BestEval Details
Integration task
# Task
Using the Daytona MCP, manage cloud sandboxes:
1. Discover available MCP tools and their capabilities
2. List all existing sandboxes (or environments/workers/deployments)...
# Task
Using the Stripe MCP, create payment resources:
1. Discover available MCP tools and their capabilities
2. Create a product called "Pro Plan" using the MCP tools
3. Create...
Generated files
1
1
Trace tool calls
5
13
Completed
9/27/2026
9/27/2026
Stderr
None captured
None captured