Continue
Continue Dev, Inc. · ai coding agents · Source: data snapshot · data as of 2026-09-16
The open-source IDE assistant layer — autocomplete, chat, and agents over any model you point it at.
ToolScore breakdown — why 77/100
10 criteria × 10 points. Low pillars are visible on purpose — that's the point of a score you can audit. Methodology →
✅ What it does well
- ▸Local-model route makes autocomplete + chat genuinely free, forever
- ▸Works in both VS Code and JetBrains — rare among OSS assistants
- ▸Config-as-code: assistant behavior is versioned YAML in the repo
- ▸No telemetry-by-default — a real privacy story
⚠️ Where it costs you
- ▸Peak quality depends entirely on the model you bring — ceiling is yours to buy
- ▸Autocomplete speed with local models lags hosted rivals on weak hardware
- ▸Extension updates have historically introduced regressions
Hidden traps — the stuff the pricing page doesn't say
The $0 path assumes a machine that can run a 7B-class model comfortably. On thin laptops, local autocomplete becomes a latency tax — budget for hosted tokens instead.
Know the limits before you pay: read the traps above, then decide with eyes open.
Visit official site →Affiliate link where available — never affects the score.
- official Continue — official site & docs (unverified snapshot) unverified
- official GitHub repo (license, config reference) unverified
Data as of 2026-09-16. Prices and limits change monthly — report anything stale. How we verify: methodology.
Closest alternatives
Open-source VS Code agent with BYOK transparency — powerful, but famous for eating tokens by design.
Google's open-source terminal agent with the most generous free tier in the category and a 1M-token
Open-source, 2nd-gen terminal agent — provider-agnostic with AST-aware context and native prompt cac
Continue review
In-depth analysis · last updated 2026-09-16
The Swiss army knife, BYO blade
Continue is less a product and more a layer: autocomplete, chat/edit, and agent modes in VS Code and JetBrains, with every model decision left to you. Point it at Ollama and your monthly bill is zero; point it at a frontier API key and it behaves like a commercial assistant with your choice of brain.
The economics of free
The local route is the headline: tab-completion that costs nothing and never leaves your machine. The trade is honest — quality and speed become functions of your hardware (6.5/10 on both for the free path). For open-source work and non-sensitive side projects, it's unbeatable; for client work on a weak laptop, hosted tokens win.
Rough edges, versioned honestly
The IDE extension has a history of update-day regressions (stability 6/10), and support is community-tier (5.5/10). What you get in exchange is config-as-code: your assistant's rules, models, and context sources live in the repo, reviewable in PRs.
Verdict
77/100, Strong. The best fit for developers who want the assistant layer open, private, and cheap — and accept a self-managed ceiling.