> A Google Docs principal engineer spent six years and $73 million rebuilding the terminal. The product pivoted three times, community trust collapsed more than once. Now it's open source — but AI does most of the work. > > This sounds like the most clichéd Silicon Valley story. But Warp's real question isn't "is it worth it" — it's: does a fifty-year-old tool like the terminal still have room for revolution?
Key points
- $73M bet: Founder Zach Lloyd (ex-principal engineer on Google Docs/Sheets) raised $23M Series A (GV, 2022) and $50M Series B (Sequoia, 2023). Angels include Sam Altman, Marc Benioff, Tobi Lütke, Jeff Weiner, Elad Gil. Sequoia bet on the terminal as the strategic entry point of developer workflow.
- Three pivots: (1) 2020–2022 Block-UI terminal — commands+output as navigable, shareable blocks; (2) 2023–2024 AI terminal — Agent Mode, Next Command, error explanation, with $20/mo Build and $50/seat/mo Business plans; (3) 2025–2026 Agentic Development Environment — Warp 2.0's Oz, a cloud platform running multiple long-lived agents (coding, testing, CI monitoring) tracked from a sidebar.
- Open source, but not fully: On April 28, 2026, Warp open-sourced under a dual license — MIT for the UI crates (
warpui_core,warpui), AGPL v3 for everything else. OpenAI is a founding sponsor; Oz runs on GPT models (including GPT-5.5). Oz itself is not open source. - "Open Agentic Development": Community issues are triaged, refined, and implemented largely by Oz agents, with humans doing final review. Hacker News critics asked whether this is community-driven development or a company automating its own engineering; others noted Warp built on Alacritty without proportionate upstream contribution. The AGPL choice is widely read as strategic: it blocks cloud vendors from forking Warp as SaaS while Warp keeps Oz closed.
- Ghostty (Mitchell Hashimoto, Zig, MIT, ~50K stars): Metal-native rendering, 120Hz, zero telemetry; moved to Hack Club 501(c)(3) nonprofit in Dec 2025.
- Alacritty (Rust, Apache 2.0): extreme minimalism, ~22MB idle memory; pair with tmux/Zellij.
- iTerm2 (macOS, GPL v2): 16 years old, feature-rich, tmux -CC integration, slowing innovation.
- Warp (Rust, AGPL+MIT, ~46K stars): blocks + AI + collaboration + agents; slowest but most featureful.
- Kitty (GPL v3): graphics protocol, kittens scripting, free and open.
- WezTerm (Rust, MIT): Lua config, built-in multiplexing, cross-platform.
- Optimistic — the "VS Code of terminals": Oz becomes an extensible agent platform, Warp Drive becomes team knowledge standard, Block UI stays the moat. $73M justified.
- Pessimistic — the "Atom of terminals": Block UI overhead never fixed, AI features outpaced by Claude Code/Copilot/GitHub tools, closed-source Oz breeds distrust, AGPL deters commercial forks. If agents converge on IDE integration, Warp's entry-point value evaporates.
github.com/warpdotdev/warp— open-source repo (AGPL + MIT, 46.1K stars)warp.dev— official site- Sacra.com profile — funding history ($23M Series A + $50M Series B = $73M)
warp.dev/blog/warp-drive-series-b— Warp Drive launch + Series B announcement- Zeniteq / ByteIota / It's FOSS — April 28–29, 2026 open-sourcing coverage
github.com/warpdotdev/warp/discussions/400— Zach Lloyd on open-source strategy (March 2026)- Ghostty 1.3.1 / Mitchell Hashimoto / Hack Club 501(c)(3)
- Hacker News
item?id=47936264— community reaction to Warp going open source
Block UI: killer feature and performance tax
Block UI adds structure to unstructured terminal output: jump between blocks, collapse confirmed ones, search within a block, share structured command+output instead of screenshots. Warp Drive builds on it — team-shared, parameterized workflows.
But the cost is real (community benchmarks):
| Metric | Ghostty | Warp | iTerm2 | |--------|---------|------|--------| | cat 100K lines | 0.7s | 1.8s | 2.4s | | cat 1M lines | 5.1s | 14.2s | 22.1s | | Input latency | ~2ms | ~8ms | ~12ms | | Idle memory (1 tab) | 28 MB | 210 MB | 85 MB | | 8 tabs after 4h | 95 MB | 380 MB | 290 MB |
Privacy and trust
From 2020–2024, Warp required login for basic terminal use — a major backlash point; the requirement was dropped in November 2024, though AI features still need an account. The client contains roughly 7,000 lines of telemetry code, on by default (opt-out available). Warp claims SOC 2 compliance and zero data retention, but the community view: closed client + default telemetry + account = unverifiable trust. Ghostty's counter-position: zero telemetry, zero accounts, zero network requests.
Competitive landscape
Minimalist camp ("the terminal should be invisible"):
Functionalist camp ("the terminal should do more"):
The philosophical split: pipe vs. workbench. Warp's predicament is trying to please both camps — minimalists find it heavy and intrusive; feature-seekers find its AI weaker than Claude Code's full-repo context.
Is $20/month worth it?
1. As a terminal: no. Ghostty is free, faster, lighter, and more privacy-respecting; Alacritty and iTerm2 serve other niches better. 2. As an AI assistant: depends. Free tier's 75 requests/month runs out fast. If you already use Claude Code or Codex, Warp's AI is redundant — it only sees terminal context, not your whole codebase. If you don't yet use an AI assistant, Warp is a low-friction entry. 3. As a team collaboration layer: possibly. Warp Drive workflows, shareable blocks, and live session collaboration are unique; valuable for 5–50 person terminal-heavy teams with complex procedures. Notion + GitHub Actions can approximate this, more generically.
Two endings
Bottom line
> Warp isn't selling a terminal — it's selling a bet: that the future entry point of developer workflow is the terminal, not the IDE. > > $20/month isn't buying a faster terminal; it's a vote on the future shape of developer workflow.