Invoke is live · free plan, paid from $20/mo · last verified 23 Oct 2026 · viability healthy
Invoke
Invoke is a desktop AI coding agent that turns Boards of dependency-linked cards into a single agent prompt, with BYOK across providers.
The dependency graph is the prompt, and that is genuinely different from pointing a chat agent at a repo and hoping. Boards plus Orchestrator mode is where multi-file leverage shows up, and five parallel agent tabs beats juggling windows. Model your spend before upgrading: code reviews consume credits and models burn them at different rates.
Verdict generatedVerified 7d ago · liveness 71/100 · cite: rightaichoice.com/tools/invoke
- Solo developers building multi-file features who want to map dependencies before prompting
- Small teams running parallel agents on independent tasks in separate tabs
- Power users who tune custom slash commands, sub-agents and thinking-effort settings
- Developers who want BYOK across providers or Ollama for local, offline-model work
- Developers who want plain chat or autocomplete with no planning step
- Anyone who wants a flat subscription without tracking a monthly credit pool
- Teams needing built-in CI/CD, deployment or hosting
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip Invoke if you want a flat-rate copilot with no credit pool to track — code reviews draw against the same monthly credits as agent runs, and different models consume them at different rates.
Code reviews consume credits on every paid tier, so a review-heavy week draws down the same pool your agents run on.
Free at $0/mo (BYOK) suits solo devs testing the Boards workflow; Starter at $20/mo is priced close to Cursor's individual tiers but adds metered credits and 10 sub-agents; Pro at $50/mo and Max at $100/mo compete with team seats on flat-rate copilots, so pick them only if 50+ sub-agents or 135k–280k monthly credits map to real work.
In short
Invoke — Invoke is a desktop AI coding agent that turns Boards of dependency-linked cards into a single agent prompt, with BYOK across providers. Best for Solo developers building multi-file features who want to map dependencies before prompting, Small teams running parallel agents on independent tasks in separate tabs, Power users who tune custom slash commands, sub-agents and thinking-effort settings. Free to start; paid plans from $20/mo.
What is Invoke?
Invoke — Invoke is a desktop AI coding agent that turns Boards of dependency-linked cards into a single agent prompt, with BYOK across providers. Best for Solo developers building multi-file features who want to map dependencies before prompting, Small teams running parallel agents on independent tasks in separate tabs, Power users who tune custom slash commands, sub-agents and thinking-effort settings. Free to start; paid plans from $20/mo.
Invoke, from Yugin Tech, is a desktop AI coding agent built around a plan-first workflow: you map features as cards on Boards, draw connections to show dependencies, then hit Build and the dependency graph itself becomes the prompt. It runs on macOS, Windows and Linux, with a CLI alongside the desktop installer. Four modes split the work — Agent for autonomous coding, Plan for scoping, Browser for web-side tasks, and Orchestrator for coordinating specialized sub-agents. A Creative Toolkit adds Canvas for drag-and-drop UI design that exports production code, and Sandbox (preview) for AI-assisted experimentation before a merge. The workspace is deliberately tool-dense: a built-in REST client with base URL management, request chaining and project API scanning, diff integration, an integrated terminal with command allow/deny controls, memories management, reusable Skills bundles, and MCP support. Recent releases added custom slash commands as markdown files in .invoke/commands/, thinking-effort controls, and on-device voice-to-text in the Agent sidebar. Bring-your-own-keys is central rather than an add-on — use your own OpenAI, Anthropic or other API keys, or run local models through Ollama. Parallel agents run up to five at once in separate tabs, and a Detached Agent window floats above your browser while you work elsewhere. Pricing is freemium on credits: Free at $0 with your own keys, Starter $20/mo for 50,000 credits, Pro $50/mo for 135,000, Max $100/mo for 280,000, with 50% rollover on paid plans. Positioning-wise, Invoke competes less with autocomplete copilots and more with chat-first agent IDEs. The bet is that a visible dependency graph produces better multi-file changes than a single long prompt.
Key features
- Boards: map features as cards, draw dependency connections, then Build the plan
- Canvas: drag-and-drop UI design that exports production code
- Sandbox (preview): AI-assisted code experimentation before merging
- Run up to five parallel agents in separate tabs
- Orchestrator Agent mode coordinating specialized sub-agents
- Custom sub-agents configured per task (3 / 10 / 50 / unlimited by tier)
- Custom slash commands created as markdown files in .invoke/commands/
- Skills: reusable knowledge bundles with instructions, scripts and references
- Thinking effort controls adjustable across AI providers
- On-device voice-to-text input in the agent sidebar (no external server)
- Built-in REST API client with endpoint testing, request chaining and API scanning
- Diff integration for code comparison and version tracking
- Git commit context menu to copy hashes or messages
- Checkpoint restore: conversation-only or conversation-and-files
- External file watching keeps the editor in sync with outside changes
Behind the verdict
We'd reach for Invoke when the work has shape — a feature that touches several files, several services, or several people's assumptions. Sketching it as cards and drawing the edges forces the dependency thinking that chat agents usually skip, and the graph then becomes the context the agent works from. If you've ever watched a coding agent confidently edit the wrong layer of an app, that planning step is the pitch. Pass on it if you want autocomplete or a single chat box. The planning step is not optional overhead you can tune away; it's the product. Developers who open an editor and want the model to just follow along will find Boards friction, not help. Browser-only and locked-down corporate machines are out too — this needs a desktop install. The closest alternative is a flat-rate agent IDE such as Cursor or Copilot. There you get set-and-forget billing and less control over which model runs where. Invoke inverts that, with your own API keys across providers plus Ollama for local work. In practice that means Invoke's value rises with how much you care about routing, cost accounting, and offline runs. Credit accounting is the real caveat. The pricing page says plainly that code reviews consume credits and that some models use more than others. A heavy review day on Starter can chew through 50,000 credits faster than the monthly framing suggests. Watch your usage panel for a couple of weeks before committing to Pro. Recent releases have been utility, not flash: rename conversations and external file watching (v0.10.23), git commit context menu and split checkpoint restore (v0.10.22), on-device voice input plus Boards box-select and node search (v0.10.21), custom slash commands and thinking-effort controls (v0.10.20). That cadence suggests a team shipping to
Models under the hood
as of 2026-10-31
Limitations
- Paid plans run on a credit system where code reviews consume credits, and Invoke's pricing FAQ states that free models carry limitations and that different models consume credits at different rates.
- The Free tier caps AI sub-agents at 3 and memory slots at 5, and requires bringing your own API keys.
- Parallel agents are documented as up to five in separate tabs, while higher sub-agent counts (10 on Starter, 50 on Pro, unlimited on Max) sit behind paid tiers.
- Sandbox is marked as preview in the docs, and on-device voice-to-text is labeled a preview feature in both the changelog and the sidebar.
- The product is a desktop application that must be installed on macOS, Windows or Linux, with no built-in CI/CD, deployment or hosting.
as of 2026-10-13
Researching Invoke? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Is Invoke still active?
Invoke is live · free plan, paid from $20/mo · last verified 23 Oct 2026 · viability healthy.
Viability score
How well maintained and how widely used is Invoke? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: October 2026
How we score →What's new in Invoke
Checked 9 days agoAcross the latest 5 updates: 4 feature updates and 1 changelog entry.
v0.10.23 Rename Conversations, External File Watching & Project Search Improvements
Rename a conversation and the tab title updates live, external file watching keeps the editor in sync with changes made outside the app, project search panel interactions improved, and a file tree fix stops stale dimensions after sleep.
v0.10.22 Git Context Menu, Checkpoint Restore Options & Bug Fixes
Copy commit hashes or messages from git history via a new context menu, and checkpoint restore is split into two options — conversation-only or conversation and files — alongside Monaco gutter and file explorer fixes.
v0.10.21 On-Device Voice Input, Boards Multi-Select & Node Search
Voice-to-text in the Agent sidebar transcribes entirely on-device with no data sent to external servers. Boards gain box-select, bulk delete, node search with keyboard navigation, and per-edge colors with inline label editing.
v0.10.20 Custom Slash Commands & Thinking Effort Controls
Create custom slash commands by adding markdown files to .invoke/commands/, adjust thinking effort across all AI providers with new textarea controls, and the 30s HTTP timeout no longer cuts off LLM streaming calls early.
v0.10.19 Bug Fixes
Resolved various stability issues and edge cases for a more reliable experience.
What do users say about Invoke?
What people actually say about Invoke — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
ResearchedDiscussed by users across 6 sources (Hacker News, YouTube, Product Hunt, App Store, Stack Overflow, Lemmy) · researched Oct 20, 2026.
Weighted by the 84 posts each of 6 sources contributed.
- +Dependency-graph Boards turn multi-file plans into a structured prompt instead of one long blob.
- +BYOK across OpenAI, Anthropic, and others means you control model cost and provider choice.
- +Ollama support lets teams keep sensitive code on local models with no cloud egress.
- +Five parallel agents in separate tabs is real throughput for larger feature work.
- +Built-in REST client with request chaining and API scanning reduces tool-switching.
- −No independent user reviews exist to validate claims about output quality or stability.
- −Name clashes with Harman Kardon Invoke and Invoke AI bury real discussion in search.
- −Credit-based pricing makes real-world monthly cost impossible to predict without benchmarks.
- −Advanced, tool-dense UI risks overwhelming users accustomed to simpler chat-first IDEs.
- −Support responsiveness is unverified — no public issue tracker or SLA visible.
- • Inference is charged to your own provider keys on top of the subscription
- • No community benchmarks to estimate whether 50,000 credits survives a week of Boards work
- • Higher tiers are partly gated on sub-agent count, nudging heavier users upward
How we verified this
Every listed fact about Invoke — pricing, features, links, status — is re-read from the vendor's own pages by our verification pipeline and each change is logged with a date; the last pass ran on 23 Oct 2026. Pages mix model-assisted synthesis with human review; see how we verify.
Verification history
We have re-verified Invoke 12 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — changed pricing tiers, features, summary, our verdict, description, who it suits, who should skip it
- — changed limitations, description, who should skip it, switching in, switching out, our verdict, features, use cases, setup time, pricing fit, hidden costs, pricing tiers, integrations, summary
- — changed who should skip it, description, summary, our verdict, who it suits, pricing tiers, features
- — changed limitations
- — changed use cases, who it suits, description, our verdict, features, hidden costs, limitations, switching in, switching out, who should skip it, pricing tiers, pricing fit, setup time, summary
- — changed who should skip it, our verdict, summary, features, description, who it suits, pricing tiers, integrations
Showing the 6 most recent of 12 verification passes.
Free to cite with attribution — this page re-verifies continuously.
How much does Invoke cost in 2026?
Invoke: Free to start; paid plans from $20/mo. Pricing last verified 23 Oct 2026.
Plans compared
Every published Invoke tier, its price as the vendor publishes it, who it fits, and what it adds over the tier below. The 12-month cost projection follows.
| Plan | Price | Billing | Who it's for |
|---|---|---|---|
| Free | Free | — | Solo developer with their own OpenAI or Anthropic keys who wants to test the Boards-to-agent workflow before paying.Adds: Starting tier: $0 with BYOK, 3 AI sub-agents and 5 memory slots, plus community Discord support. |
| Starter | $20/mo | monthly | Individual developer running daily agent work on personal projects who needs Orchestrator mode and more memory.Adds: Adds 50,000 AI credits/mo, Orchestrator Agent mode, 10 sub-agents, 15 memory slots and 50% credit rollover. |
| Pro | $50/mo | monthly | Power user or small team automating complex multi-file work who would otherwise hit the Starter credit ceiling.Adds: Raises the pool to 135,000 credits/mo, 50 sub-agents and 25 memory slots, with faster priority support. |
| Max | $100/mo | monthly | Serious builders and teams running heavy parallel agent workloads where sub-agent count is the bottleneck.Adds: 280,000 credits/mo, unlimited AI sub-agents and 50 memory slots, plus dedicated priority support. |
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Where the pricing makes sense
The company stage and team size where Invoke's pricing actually pencils out — and where peers do it cheaper.
Free at $0/mo (BYOK) suits solo devs testing the Boards workflow; Starter at $20/mo is priced close to Cursor's individual tiers but adds metered credits and 10 sub-agents; Pro at $50/mo and Max at $100/mo compete with team seats on flat-rate copilots, so pick them only if 50+ sub-agents or 135k–280k monthly credits map to real work.
Has Invoke's pricing changed?
No plan price for Invoke has moved since we began logging in Aug 2026; the pricing page is re-read on every verification pass.
Who should pick Invoke?
Pick Invoke if you are Solo developers building multi-file features who want to map dependencies before prompting, Small teams running parallel agents on independent tasks in separate tabs, Power users who tune custom slash commands, sub-agents and thinking-effort settings, Developers who want BYOK across providers or Ollama for local, offline-model work.
Look elsewhere if you are Developers who want plain chat or autocomplete with no planning step, Anyone who wants a flat subscription without tracking a monthly credit pool, Teams needing built-in CI/CD, deployment or hosting, Browser-only or locked-down environments where a desktop install isn't possible.
Skip it if Skip Invoke if you want a flat-rate copilot with no credit pool to track — code reviews draw against the same monthly credits as agent runs, and different models consume them at different rates.
GeneratedReal-world workflow fit
Concrete scenarios for the personas Invoke actually fits — and what changes day-one when you adopt it.
You open a Board, add cards for each file or component the feature touches, and draw edges to show which changes depend on which. Press Build and Orchestrator mode dispatches sub-agents in dependency order while you watch the diff pane.
Outcome: The agent works in the order you specified instead of guessing, so you review a coherent set of changes rather than a pile of unrelated edits.
One dev keeps a Detached Agent window floating over a browser for web-side tasks while another runs independent agents in separate tabs, each scoped to a card on the shared Board.
Outcome: Up to five agents progress in parallel without stepping on each other, and git commit context menus plus checkpoint restore give you a clean rollback line if one branch goes wrong.
You point Invoke at Ollama for local models, keep terminal command allow/deny rules tight, and use on-device voice-to-text in the Agent sidebar to dictate scope without audio leaving the machine.
Outcome: Planning and code generation stay on-device, and the only metered cost is whatever local inference uses.
Use cases
- Plan a complex feature set on Boards, draw the dependency edges, and let agents build them in order.
- Design UI pages visually in Canvas and export production code.
- Run up to five independent agents in separate tabs on parallel tasks.
- Define custom slash commands and sub-agents for repetitive code review, refactoring or documentation work.
- Test and chain REST endpoints and scan your project for APIs without leaving the editor.
- Experiment with risky changes in Sandbox and merge with AI assistance.
- Keep work offline and on-device by routing agents through Ollama with your own keys.
- Unify diff review, git history, terminal and API testing inside one desktop workspace.
Setup time & first value
How long it actually takes to get something useful out of Invoke — broken out by persona, not the marketing-page minute.
Install the desktop build on macOS, Windows or Linux and you're in the editor in a few minutes. Add your provider keys (or point at Ollama) and most developers reach first useful agent output inside an hour. Getting real leverage from Boards, Skills and custom slash commands takes a day or two of structuring your own workflow around them.
Switching to or from Invoke
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From Cursor: keep your repo as-is, point Invoke at the same provider keys, and move planning out of chat into Boards so dependency order is explicit.
- →From GitHub Copilot: add Invoke alongside for multi-file agent runs and bring your BYOK keys, since Invoke's Free tier uses your own API access.
- →From a chat-only assistant: recreate your recurring review and refactor prompts as markdown slash commands in .invoke/commands/ so they run in one keystroke.
- →From a local-only setup: keep your Ollama models and add Invoke's Board, diff and REST client surface around them.
- ↗To Cursor or GitHub Copilot: export your code as normal and accept losing the Board dependency graph — flat-rate billing replaces the credit meter.
- ↗To a chat-only assistant: copy your Skills and slash-command markdown out of .invoke/commands/ and re-paste the prompts by hand.
- ↗To a CI-integrated platform: move code review and deployment checks to a tool with built-in CI/CD, since Invoke covers authoring rather than shipping.
Integrations
Resources & guides
- Documentationinvoke.yugin.co
Docs · Invoke
Full product docs from invoke.yugin.co
- Documentationinvoke.yugin.co
Installation · Invoke
Full product docs from invoke.yugin.co
- Documentationinvoke.yugin.co
Models · Invoke
Full product docs from invoke.yugin.co
- Documentationinvoke.yugin.co
Skills · Invoke
Full product docs from invoke.yugin.co
- Resourceinvoke.yugin.co
Changelog · Invoke
Helpful link from invoke.yugin.co
Tutorials & Learning
YouTube returned 6 videos for “Invoke”, and we withheld 6: 6 could not be judged, because “Invoke” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about Invoke.
Official links
What are the best Invoke alternatives?
The closest alternatives to Invoke we list are GitHub Copilot, Cursor and Design.MD. Ranked by direct product-type match, not category overlap.
| Alternative | Pricing | Starts at | Last verified |
|---|---|---|---|
| GitHub Copilot | Freemium | Pricing not published | — |
| Cursor | Freemium | Pricing not published | — |
| Design.MD | Freemium | Pricing not published | — |
Alternatives to Invoke
View allGitHub Copilot
GitHub's AI coding agent for inline completion, delegated pull requests, code review, and recurring agent tasks across your repos.
Tools that pair well with Invoke
Common stack mates teams adopt alongside Invoke, with the specific reason each pairing earns its keep.
GitHub Copilot
GitHub's AI coding agent for inline completion, delegated pull requests, code review, and recurring agent tasks across your repos.
Cursor
Cursor is an AI coding agent that plans, builds and ships features from your editor, CLI, Slack, GitHub and iOS.
Design.MD
DESIGN.md files that give your AI coding agent a real brand's colors, type and spacing instead of generic AI layouts.
Invoke vs its closest rival
Invoke's closest head-to-head is with Polycam.
Our Invoke vs Polycam verdict written Jul 3, 2026 is more than 90 days old, so it is being re-reviewed against current data and is not shown.
Frequently asked questions
Best-of guides
Used Invoke? Help shape our editorial sentiment research.