Amp Inc.'s CLI coding agent — independent spin-out from Sourcegraph (March 2026). Sub-agent architecture (Oracle, Librarian, Painter) built on Sourcegraph code intelligence DNA. 36,981 npm weekly downloads. Free tier + BYOK. Note: original sourcegraph/amp GitHub repo now 404s.
Where it wins
Sub-agent architecture: Oracle (code analysis), Librarian (external libraries), Painter (code review) — most sophisticated in category
Sourcegraph code intelligence DNA — deep codebase search and code graph
Available across terminal, VS Code, Cursor, Windsurf, JetBrains, Neovim
Free tier + BYOK/pay-as-you-go — pricing improved from $59/user/mo
Deep Mode for autonomous research + extended reasoning
36,981 npm weekly downloads — real adoption signal
Where to be skeptical
Closed source — all threads stored on Sourcegraph servers, independent verification impossible
No published SWE-bench Pro or Terminal-Bench benchmark scores
CLI-only pivot (editor extension self-destructed March 5) — unproven at scale
Real switching stories are individual anecdotes, not systematic evidence
HN engagement (90 pts) is 24x lower than Claude Code
Editorial verdict
Watch — corporate spin-out from Sourcegraph to independent Amp Inc. (March 2026) is a material change. The original sourcegraph/amp GitHub link returns 404. Tool still ships (36K npm downloads/week) under ampcode.com. Sub-agent architecture (Oracle, Librarian, Painter) remains the most sophisticated in the category. Update or verify all links before recommending.
Videos
Reviews, tutorials, and comparisons from the community.
How to Build Vertical AI Businesses Fast with Ryan Carson, Builder in Residence at Sourcegraph
GPT 5.4 LIVE Test & Learn to Code in 2026: What's Essential vs. What AI Handles Now
Related
Coding CLIs / Code Agents
Teams with large, complex codebases needing deep code intelligence
Software Factories
CLI-first power users who value Sourcegraph code intelligence lineage and BYOK flexibility
Claude Code
98Anthropic's official agentic coding CLI. v2.1.81 (Mar 20) shipped `--bare`, smarter worktree resume, and improved MCP OAuth while the repo crossed 82,204 stars and logged ~14 commits/week across 10+ maintainers. Terminal-native, tool-use-driven, with deep file system + shell access, #1 SWE-bench Pro standardized (45.89%), ~4% of GitHub public commits (SemiAnalysis), $2.5B annualized revenue. 8M+ npm weekly downloads. Opus 4.6 with 1M context.
OpenHands
88Category leader in multi-agent orchestration — 69,352 stars (verified), $18.8M Series A, AMD hardware partnership, 455 contributors, 1M downloads/month PyPI (3.4M all-time). SWE-Bench Verified 72% with Claude 4.5 Extended Thinking (updated 2026-03-19), Multi-SWE-Bench #1 across 8 languages. Gap to #2 is enormous on every axis.
OpenCode
88Open-source AI coding agent from SST. v1.2.27 active (2026-03-16) — development resumed after a gap. OpenAI official partnership following the Anthropic OAuth block controversy. 126K+ GitHub stars (star surge driven by Anthropic controversy). Known unauthenticated RCE fixed in v1.1.10+ (CVE, 432 HN pts). CVE-2026-22812 (CVSS 8.8-10.0) is a second serious security incident.
Gemini CLI
88Google's open-source terminal agent with Gemini 3 models, 1M token context, built-in Google Search grounding, and the best free tier in the category (60 req/min, 1K req/day). v0.35.0 (Mar 24) shipped keybinding, policy, and telemetry fixes while the repo hit 98,957 stars and 12,593 forks. Terminal-Bench 2.0: 78.4% (#1). SWE-bench Pro standardized 43.30% (#3). Plan Mode added March 2026. First-pass correctness ~50-60% (Educative.io).
Public evidence
Individual switching story from Cursor to Amp. Praises deep code intelligence and sub-agent architecture. Anecdotal, not systematic evidence.
Strategic bet on CLI-only. Sub-agents (Oracle, Librarian, Painter) for deep code intelligence. Free tier and BYOK/pay-as-you-go pricing replaces $59/user/mo.
Raw GitHub source
GitHub README could not be fetched right now.