Overall in coding agents
2of 17 · score 82.4
Data coverage
62%
16 of 26 fields sourced
Best for
- •Nothing we can responsibly recommend yet
Strong on
- +Permissions allowlist + OS sandbox 1/17
- +Model breadth any provider / router 1/17
- +Subagents Yes 1/16
Weak on
- −GitHub stars 28,000 7/11Better: Claude Code: 155,000
- −Surface CLI 13/17Better: GitHub Copilot (Agent HQ): GitHub-native, IDE extension, CLI, web, cloud delegate
vs GitHub Copilot (Agent HQ)
- •Ahead on Licence: Apache-2.0 vs proprietary
- •Ahead on BYOK: Yes vs Partial
- •Behind on Surface: CLI vs GitHub-native, IDE extension, CLI, web, cloud delegate
How Grok Build scores among 17 coding agents16 of 26 fields known
Features8
Surface
CLI
13/17
Leader: GitHub Copilot (Agent HQ) · GitHub-native, IDE extension, CLI, web, cloud delegate
x.ai · 2026-11-07
Repo indexing
—
Checkpoint & undo
—
Agentic7
Governance & control3
Pricing3
Pricing model
—
Entry price
—
Token markup
—
Traction3
Terminal-Bench
—
Bench model
—
Positioning1
Positioning
A powerful, extensible coding agent used via an interactive TUI, headlessly in scripts or bots, or through ACP in other apps.
docs.x.ai · 2026-11-07
Timeline1
Alternatives8
Aider
Amp
Claude Code
Cline
Codex CLI
Cursor
Devin
DIY harnessCompare
Sources7
- x.ai/news/grok-build-cli8 fields
- raw.githubusercontent.com/xai-org/grok-build/main/crates/codegen/xai-grok-pager/docs/user-3 fields
- docs.x.ai/build/features/sandbox1 field
- raw.githubusercontent.com/xai-org/grok-build/main/crates/codegen/xai-grok-pager/docs/user-1 field
- raw.githubusercontent.com/xai-org/grok-build/main/README.md1 field
- api.github.com/repos/xai-org/grok-build1 field
- docs.x.ai/build/overview1 field