Coding agents · CA3
Turn a GitHub issue into a PR autonomously
9 candidates. None have been tested, so this list is alphabetical, not a rank.
Standard test task
Assign 3 seeded issues (bug, small feature, docs) in a sample repo; judge merged-ready PRs without human edits.
Not tested yet. A dash is an empty score, not a zero. We do not estimate results, prices, or ranks.
Not tested
Amp
Pay per use (credits); bring-your-own subscription/keys supported
No free tier
Checked 15 Oct 2026
Amp Free closed to new users since Feb 2026; existing free allowance reduced May 2026.
- Job resultWeight 40
- —
- Speed and effortWeight 15
- —
- Price and free tierWeight 15
- —
- Ease of startWeight 10
- —
- Reliability and trustWeight 10
- —
- Community signalWeight 10
- —
- Last tested
- Not tested yet
- Evidence
- No test evidence yet.
No score. We do not estimate one.
Not tested
Claude Code
Anthropic's terminal/IDE coding agent
- Job resultWeight 40
- —
- Speed and effortWeight 15
- —
- Price and free tierWeight 15
- —
- Ease of startWeight 10
- —
- Reliability and trustWeight 10
- —
- Community signalWeight 10
- —
- Last tested
- Not tested yet
- Evidence
- No test evidence yet.
No score. We do not estimate one.
Not tested
Cursor
AI-first code editor with agent mode and background agents
- Job resultWeight 40
- —
- Speed and effortWeight 15
- —
- Price and free tierWeight 15
- —
- Ease of startWeight 10
- —
- Reliability and trustWeight 10
- —
- Community signalWeight 10
- —
- Last tested
- Not tested yet
- Evidence
- No test evidence yet.
No score. We do not estimate one.
Not tested
Devin
Free $0 (light quota); Pro $20/mo; higher tiers
Free tier
Checked 15 Oct 2026 · Source (opens in a new tab)
Cognition's autonomous software engineer
- Job resultWeight 40
- —
- Speed and effortWeight 15
- —
- Price and free tierWeight 15
- —
- Ease of startWeight 10
- —
- Reliability and trustWeight 10
- —
- Community signalWeight 10
- —
- Last tested
- Not tested yet
- Evidence
- No test evidence yet.
No score. We do not estimate one.
Not tested
Factory
Paid with trial
No free tier
Pricing being checked
Why this price is unverified
unverified (pricing page JS-rendered) https://factory.ai/pricing
'Droids' agents for end-to-end dev tasks
- Job resultWeight 40
- —
- Speed and effortWeight 15
- —
- Price and free tierWeight 15
- —
- Ease of startWeight 10
- —
- Reliability and trustWeight 10
- —
- Community signalWeight 10
- —
- Last tested
- Not tested yet
- Evidence
- No test evidence yet.
No score. We do not estimate one.
Not tested
GitHub Copilot
Completions, chat, coding agent and PR review in GitHub/IDEs
- Job resultWeight 40
- —
- Speed and effortWeight 15
- —
- Price and free tierWeight 15
- —
- Ease of startWeight 10
- —
- Reliability and trustWeight 10
- —
- Community signalWeight 10
- —
- Last tested
- Not tested yet
- Evidence
- No test evidence yet.
No score. We do not estimate one.
Not tested
Jules
Free tier + paid
Free tier
Pricing being checked
Why this price is unverified
unverified (pricing page not fetchable: bot-blocked/JS/no /pricing)
Google's async coding agent for GitHub repos
- Job resultWeight 40
- —
- Speed and effortWeight 15
- —
- Price and free tierWeight 15
- —
- Ease of startWeight 10
- —
- Reliability and trustWeight 10
- —
- Community signalWeight 10
- —
- Last tested
- Not tested yet
- Evidence
- No test evidence yet.
No score. We do not estimate one.
Not tested
OpenAI Codex
Included in ChatGPT plans / API
No free tier
Pricing being checked
Why this price is unverified
unverified (pricing page not fetchable: bot-blocked/JS/no /pricing)
OpenAI's cloud and CLI coding agent
- Job resultWeight 40
- —
- Speed and effortWeight 15
- —
- Price and free tierWeight 15
- —
- Ease of startWeight 10
- —
- Reliability and trustWeight 10
- —
- Community signalWeight 10
- —
- Last tested
- Not tested yet
- Evidence
- No test evidence yet.
No score. We do not estimate one.
Not tested
OpenHands
Open-source autonomous coding agent
- Job resultWeight 40
- —
- Speed and effortWeight 15
- —
- Price and free tierWeight 15
- —
- Ease of startWeight 10
- —
- Reliability and trustWeight 10
- —
- Community signalWeight 10
- —
- Last tested
- Not tested yet
- Evidence
- No test evidence yet.
No score. We do not estimate one.
Weights and rules are on How we rank.