Agentic Index
Google Jules vs Capy (2026)
Both vendors last verified
Google Jules is winding down while Capy keeps shipping, so the comparison now turns on where Jules work goes next. That verdict is the Agentic Index coverage score, graded from each vendor's own published materials.
Google stopped new Jules sessions on 11 November 2026 and shuts the service on 31 December 2026, with an export of settings, scheduled tasks, memories and session history through the Jules Tools CLI. Capy, the agent platform from Scrapybara, Inc., covers the same ground: it takes an issue, works in its own cloud machine and opens a pull request.
Both ran each task in an isolated cloud VM, worked from GitHub, checked the running app in a browser and offered a free way in. Jules ran only on Google's Gemini models for individual Google accounts, with Pro at $19.99 a month through Google AI Pro. Capy lets a team pick the model for each thread from OpenAI, Anthropic, Google, xAI and others, runs agents in its cloud, on a laptop or on any server, and charges one organization price from Lite at $20 a month with no seat fee.
Choose Jules only to read or export past sessions before the shutdown. Choose Capy to keep handing issues to an agent that opens pull requests.
On the Agentic Index coding agent ranking, neither Google Jules nor Capy clears the bar, which asks for all five merge loop capabilities documented in full. Neither documents knowledge grounding and RAG in full, nor testing, debugging and optimization. 3 of the 96 vendors in the lane clear it. See the coding agent ranking
This comparison is published by Agentic Index, an independent agentic AI vendor research platform. Google Jules and Capy are each graded against the same 14 capability Agentic Index taxonomy, from the vendor's own public materials under the Agentic Index verification standard, alongside 1,212 researched vendors. No vendor pays for placement and no vendor has reviewed this page. How this evidence is graded
Choose Google Jules if
- You have past Jules work to keep. Until 31 December the Jules Tools CLI exports account settings, environment variables, scheduled tasks, repositories, memories, setup scripts and session history.
- You only need to look back. Jules stays read only until the shutdown, and the code and pull requests it opened stay on GitHub.
- Your team will follow Google's next tool. Google is folding what it learned from Jules into Antigravity, its agentic development platform.
- You want a record of how Jules worked before moving. It planned each task for approval, ran tests in its VM, fixed failing CI checks on its own pull requests and worked on Gemini models.
Choose Capy if
- You need a home for the issues Jules used to take. Capy picks up Linear issues, @capy mentions on pull requests and Slack threads, works each in its own Firecracker microVM, and never merges unless asked.
- You want more than Gemini. Each thread and subagent can run on models from OpenAI, Anthropic, Google, xAI, Moonshot, DeepSeek and others, or on your own Codex, Copilot or SuperGrok subscription.
- Big jobs need many agents. Captains orchestrate hundreds of agents, splitting a migration into subagents that each ship their own pull request.
- A team will share it. One organization subscription covers unlimited members with no seat fee, from Lite at $20 a month to Max at $200, with an audit log on every plan and SSO as an add on.
- Agents should run where your code lives. Threads run in Capy's cloud, on a Mac, Windows or Linux computer through the desktop app, or on any server through the Dock CLI.
| Feature | G Google Jules |
C Capy |
|---|---|---|
| Action & orchestration | ||
|
Integrations & Tool Calling Ability to connect agents to real systems through native integrations, OAuth-authenticated actions, custom tools, APIs, webhooks, or MCP-compatible tools. |
||
|
Google JulesIntegrations & Tool Calling Deep GitHub integration covers issue labels, branches, pull requests and reading and responding to PR comments, MCP servers connect Linear, Stitch, Neon, Tinybird, Context7 and Supabase by API key in Settings, a Render integration detects failed builds and pushes fixes, and the REST API wires Jules into Slack, Linear, Jira and CI pipelines including GitHub Actions. GitHub is the only source control host. The agent runs shell commands, installs dependencies and executes builds inside its VM with repository level environment variables. Sourcejules.google/docs/changelog 2026-02-02 and 2025-10-03, docs/integrations/renderread 2026-10-28 |
||
|
CapyIntegrations & Tool Calling Capy acts inside the tools a team uses. A GitHub App opens and reviews pull requests, Slack threads take @Capy, delegating a Linear issue creates a native agent session that streams progress back, and Sentry reports can be imported. MCP servers connect over HTTP or stdio with OAuth, Tailscale ephemeral nodes reach private databases and internal services, Vercel previews, Revyl and Paper are integrated, and OIDC federation gives machines short lived GCP, AWS, Azure or Vault credentials. GitHub Enterprise Server is not supported. Sourcedocs.capy.ai/integrations/mcpread 2026-10-28 |
||
|
Workflow Orchestration Ability to sequence, branch, retry, route, and combine deterministic workflow nodes with autonomous agent steps. |
||
|
Google JulesWorkflow Orchestration A task runs as a multi step plan the agent produces before writing code, with an Interactive Plan mode where it asks clarifying questions and refines the approach collaboratively, a Critic agent performing adversarial review of every proposed change during generation, a Planning Critic reviewing auto approved plans and reported to cut task failure rates by 9.5 percent, a CI Fixer loop that detects failures, fixes, commits and resubmits, and up to five parallel sessions on one prompt via the CLI --parallel flag, within the plan's concurrency limit of 3, 15 or 60 tasks. Sourcejules.google/docs/changelog 2026-01-26-1, 2026-02-19 and 2025-11-10, docs/cli/reference and docs/usage-limitsread 2026-10-28 |
||
|
CapyWorkflow Orchestration A thread delegates work to subagents that nest up to three levels, on a shared or a fresh machine, each with its own model, and work such as a migration splits across subagents that each ship their own pull request. Captains are long lived orchestrator agents that direct crews of threads, keep their own Drive and can manage hundreds of agents, so a team can run thousands of agents at once, each task on its own branch and machine. Sourcedocs.capy.ai/subagentsread 2026-11-19 |
||
|
Triggers & Channel Coverage How agents wake up and where they work: schedules, webhooks, message events, CRM events, inbox events, chat, email, voice, and collaboration tools. |
||
|
Google JulesTriggers & Channel Coverage Tasks start from a jules label on a GitHub issue, the web app, the Jules Tools CLI, a REST API for CI/CD and chat integrations, scheduled recurring tasks (daily, weekly or monthly) with pause and resume, proactive suggested tasks where Jules scans a repository for TODO comments and surfaces work, pull request comments which the agent reads and acts on, and automatic triggers on CI failures in GitHub Actions and failed Render builds. No inbound webhook is documented. Sourcejules.google/docs/guides/scheduled-tasks, docs/guides/running-tasks and docs/changelog 2025-12-10, 2026-01-26 and 2026-02-19; jules.google/docs/changelog/2025-12-101read 2026-10-28 |
||
|
CapyTriggers & Channel Coverage Automations carry one to twenty triggers: cron with a five minute floor, GitHub events, Slack message bursts, an incoming webhook and on demand runs from the API, and the API adds Linear triggers (issue created, status changed, end of cycle) and Sentry triggers. Delegating a Linear issue or mentioning @capy on a pull request comment starts work, and a thread wakes on its own when CI finishes, a review lands or a pull request merges. Sourcedocs.capy.ai/automationsread 2026-10-28 |
||
| Knowledge & context | ||
|
Knowledge Grounding & RAG Ability to ground agent behavior in company data through document ingestion, retrieval, external knowledge APIs, semantic search, or RAG layers. |
||
|
Google JulesKnowledge Grounding & RAG Each task clones the repository into its VM and reads it directly, with AGENTS.md conventions read from the repository root, a file selector to scope context, image upload and web search for current documentation. Suggested Tasks scan up to five repositories for TODO comments. Context is assembled per task, and the docs describe no persistent index, graph or embeddings layer over the customer's code. Sourcejules.google/docs, docs/guides/suggested-tasks and docs/environmentread 2026-10-28 |
||
|
CapyKnowledge Grounding & RAG Agents work over a real checkout of the customer's repository inside their machine, nested AGENTS.md files load when a folder is first touched, a Review context page holds organization rules and files, and imported context such as Sentry reports and attached files lands on the agent's machine. No persistent codebase index, embeddings or retrieval layer is described, so what an agent knows of the code comes from reading the checkout during the task. Sourcedocs.capy.ai/instructionsread 2026-10-28 |
||
|
Memory & State Persistence Ability to persist context across a run, conversation, workflow, user, team, or longer-term memory layer. |
||
|
Google JulesMemory & State Persistence Repository Memory saves the user's preferences, nudges and corrections during a task and references them on later tasks in that repository, toggled on or off per repository under Knowledge. The changelog is the only page that describes it. Memory has no stated lifetime and no documented way to review or delete entries, and environment snapshots and variables are setup state, not memory. Sourcejules.google/docs/changelog 2025-09-30read 2026-10-28 |
||
|
CapyMemory & State Persistence A thread is a durable conversation stored on Capy's servers that holds the full history of its messages, machines, subagents and pull requests, survives archiving and resumes when woken. An opt in Memory experiment keeps what a user teaches Capy across threads, and review findings list the memory they rest on, but no scope, lifetime or way to delete entries is described. Drive holds AGENTS.md instructions at personal or organization level, which are instructions, not memory. Sourcecapy.ai/changelogread 2026-10-28 |
||
| Control & trust | ||
|
Human Oversight & Guardrails Approval steps, consent checkpoints, escalation rules, structured guardrails, policy constraints, and pause/resume controls. |
||
|
Google JulesHuman Oversight & Guardrails The agent writes a step by step plan the user approves, edits or rejects before any code is written, with an Interactive Plan mode asking clarifying questions first, and the user can chat with it at any point. If the user leaves, the plan auto approves on a timer, and the API's requirePlanApproval option is off by default, so a Planning Critic reviews plans that are auto approved. Work lands on a branch and the user chooses when to publish or open a pull request, a Critic agent adversarially reviews each proposed change during generation, tasks can be paused, resumed or deleted mid run, and commit authorship is configurable so changes are attributed to Jules, co-authored, or the user. Sourcejules.google/docs/guides/review-plan, docs/api/reference and docs/changelog 2026-01-26-1 and 2026-02-19; jules.google/docs/changelog/2025-08-082read 2026-10-28 |
||
|
CapyHuman Oversight & Guardrails Capy never merges or enables auto merge unless asked in the thread, the Merge menu offers Enable auto merge and Force merge, and a root AGENTS.md can grant standing merge permission. The agent waits for a go ahead when told to, review mode runs Off, Once or on every push, paired devices ask before use, any agent can be stopped mid run, and a thread or Captain tags the user under Needs you when blocked on a decision, access or credentials. Sourcedocs.capy.ai/admin/securityread 2026-10-28 |
||
|
Security, Identity & Governance RBAC, SSO, auditability, encryption, least-privilege tool access, compliance posture, and data handling policy. |
||
|
Google JulesSecurity, Identity & Governance Jules isolates each task in a cloud VM, reaches repositories only as granted through the GitHub app, stores integration API keys encrypted, does not train on private repositories and tells users they are responsible for the code they run. There is no access surface (SSO, roles, organization administration) and no attestation or trust page, and paid plans are sold to individual Google accounts only. Sourcejules.google/docs/faq and docs/usage-limitsread 2026-10-28 |
||
|
CapySecurity, Identity & Governance Capy holds SOC 2 Type II on every plan, audited and pen tested every year, with the report available through its trust center. Organizations carry Owner, Admin and Member roles, and every plan includes an audit log. Single sign on through SAML and OIDC with Okta, Google Workspace and Microsoft Entra ID, with required MFA, is a $100 a month add on for Lite, Pro and Max and included on Enterprise, and the desktop app signs in with SSO. Zero data retention is available on request, code is not used for training, and data is encrypted in transit and at rest. No SCIM is documented. Sourcecapy.ai/pricingread 2026-11-19 |
||
|
Observability & Auditability Traces, logs, execution histories, metrics, audit events, and debugging detail for production agent behavior. |
||
|
Google JulesObservability & Auditability Each session carries a collection of activities representing the steps the agent took. Google calls them immutable, following an event source pattern and usable to reconstruct the session, and they are retrievable through the REST API with a createTime range cursor. The web app shows a real time activity feed, per task logs and task history per repository, the full changeset is downloadable in git patch format, and the critic agent's step by step evaluation is visible in the UI as it runs. No organization audit log is documented. Sourcejules.google/docs/changelog 2026-01-26-4 and 2025-09-03, docs/guides/coderead 2026-10-28 |
||
|
CapyObservability & Auditability Each thread keeps its full transcript and tool output, shows the agent's plan, terminal output, file changes and diffs, and a status (active, waiting, needs attention, ready for review), with the machine's desktop streamed live into the thread pane. An Activity page and usage reports break spend down by member, model and thread, the API streams events, end to end test runs record annotated video, and every plan includes an audit log. Sourcedocs.capy.ai/async-workread 2026-11-19 |
||
|
Deployment & Data Residency Deployment modes and options, including SaaS, dedicated cloud, VPC, on-prem, hybrid, local runtime, and self-hosting. |
||
|
Google JulesDeployment & Data Residency Every task runs in a fresh, short lived Ubuntu VM in Google's cloud with internet access, whose toolchain the customer shapes with a setup script, snapshots and environment variables. That configures the environment, not where it runs. There is no region choice, customer VPC, on premises or tenant option, so a customer with a different residency requirement has no other answer. Sourcejules.google/docs/environment and docs/faqread 2026-10-28 |
||
|
CapyDeployment & Data Residency Each cloud thread runs on its own Firecracker microVM on Ubuntu 24.04, sleeping after about two minutes and deleted after 30 days asleep. The hosted service runs in one region, US East (New York), with no region choice. Threads can also run on the customer's own Mac, Windows or Linux computer through the desktop app or on any server through the Dock CLI, moving between machines mid task, and Enterprise offers running in the customer's own cloud on request, with OIDC federation issuing short lived cloud credentials to machines. Sourcedocs.capy.ai/machinesread 2026-11-19 |
||
| Solution readiness | ||
|
Prebuilt Agents, Templates & Packs Ready-made workflows, packaged employees, templates, blueprints, industry solutions, and role-specific agents that reduce time-to-value. |
||
|
Google JulesPrebuilt Agents, Templates & Packs Sample prompts ship as one click starting points, three scheduled task templates (Performance, Design and Security) start recurring tasks, suggested tasks let the agent propose work it has found in the repository, and MCP integrations are offered from a vendor curated list of vetted servers. There is no library of prebuilt agents, installable skills or marketplace, and the community prompt list lives in a GitHub repository. Sourcejules.google/docs/guides/scheduled-tasks, docs and docs/changelog 2025-09-02 and 2026-02-02; jules.google/docs/changelog/2025-09-02read 2026-10-28 |
||
|
CapyPrebuilt Agents, Templates & Packs A plugin marketplace offers searchable, categorized plugins that bundle MCP servers and skills, with a Verified by Capy mark on plugins Capy has picked and tested, and named plugins include Firecrawl, ClickStack and Supermemory. A team installs plugins across the organization or per project or adds one from a GitHub repository. An Explore page lists plugins, skills and themes, and built in skills ship with the product beside the Review agent and Captains. Sourcedocs.capy.ai/integrations/pluginsread 2026-10-28 |
||
| Platform extensibility | ||
|
Model Flexibility & Routing Ability to work across multiple foundation models, route tasks to different models, or let buyers bring their own providers and keys. |
||
|
Google JulesModel Flexibility & Routing Every model Jules runs is Google's Gemini. The changelog makes Gemini 3 Flash the base model on all tiers and Gemini 3.1 Pro the default for Pro subscribers, which is a tier choice within one maker, while the homepage and usage page still name Gemini 2.5 Pro and Gemini 3 Pro. There is no user model choice, no other maker and no bring your own model path. Sourcejules.google/docs/changelog 2026-03-09 and 2026-01-30, docs/usage-limitsread 2026-10-28 |
||
|
CapyModel Flexibility & Routing A person picks the model per thread and per subagent from Anthropic, OpenAI, xAI, Google, Moonshot, Z.ai, DeepSeek, Alibaba, MiniMax, Meta and other makers, and can run one task on several models to compare. Teams can bring Anthropic, OpenAI, Google and xAI keys or gateways such as Vercel, OpenRouter and Cloudflare, and existing Codex, Copilot and SuperGrok subscriptions connect and draw on the provider's limits. Model tokens bill at each provider's list price with no markup. Sourcedocs.capy.ai/models-and-pricingread 2026-10-28 |
||
|
APIs, SDKs & MCP Extensibility Composability layer: stable APIs, SDKs, MCP tool consumption/serving, custom tools, and integration into internal systems. |
||
|
Google JulesAPIs, SDKs & MCP Extensibility A REST API at jules.googleapis.com/v1alpha, in alpha, authenticates with an X-Goog-Api-Key header (up to three keys) and exposes Sources, Sessions (create, list, approve plan, send message, auto create PR, repoless sessions) and Activities with git patch output. A Jules Tools CLI installs via npm as @google/jules with jules remote new, list and pull, parallel execution and local patch application, a Gemini CLI extension exposes Jules as a slash command, and MCP server support connects a vetted set of servers by API key. Sourcejules.google/docs/api/reference, docs/api/reference/authentication, docs/cli/reference and docs/changelog 2026-02-02; jules.google/docs/changelog/2026-02-02read 2026-10-28 |
||
|
CapyAPIs, SDKs & MCP Extensibility A public REST API with bearer keys and an OpenAPI specification covers projects, threads, messages, tasks, automations, reviews and usage, so developers can start, message, interrupt and archive agent work from their own code and stream its events. MCP servers and plugins extend what the agents can reach, and the Dock CLI attaches any server as a machine for Capy threads. Sourcedocs.capy.ai/api-reference/overviewread 2026-10-25 |
||
|
Testing, Debugging & Optimization Testing, debugging, scoring, retries, fallbacks, quality gates, and optimization loops for improving agent workflows before and after deployment. |
||
|
Google JulesTesting, Debugging & Optimization The agent runs the project's existing test suite and iterates when tests fail, writes tests on its own, and the CI Fixer detects a failed GitHub Actions check, works through a fix and resubmits the pull request in a loop. A Critic agent adversarially reviews proposed changes and a Planning Critic reviews plans, both internal quality mechanisms, and Run to Validate checks environment setup scripts. These verify the customer's code and the agent's own output. There is no harness for customers to evaluate or regression test agent behavior. Sourcejules.google/docs/changelog 2026-02-19 and 2026-01-26-1, docs/environmentread 2026-10-28 |
||
|
CapyTesting, Debugging & Optimization A built in Review agent checks every pull request from a real checkout, labels each finding confirmed or investigate and comments on specific lines, and Autofix answers CI failures and review comments. Agents run the project's own tests and end to end checks in their machines, which tests the customer's code. On the agent side, the same task can run on different models to compare approaches, and there is no evaluation suite or regression testing of agent behavior. Sourcedocs.capy.ai/reviewread 2026-10-28 |
||
| Specialist automation | ||
|
Browser & Computer Use Browser, desktop, or remote/local computer control for workflows that cannot be handled through stable APIs alone. |
||
|
Google JulesBrowser & Computer Use The default Jules base image includes Playwright and ChromeDriver for front end testing, and the agent renders a running web application and returns a screenshot of the front end alongside its code changes when asked to verify its work, with images shown in the diff viewer. With Gemini 3 Pro, visual verification of web app outcomes is more precise, and Jules can also search the web for documentation and code examples during a task. Sourcejules.google/docs/changelog 2025-08-07 and 2025-11-19, docs/environmentread 2026-10-28 |
||
|
CapyBrowser & Computer Use Each cloud machine runs an XFCE desktop with Google Chrome that streams live into the thread pane, and the agents use native computer use and browser use for end to end testing, recording annotated videos of each run. From the Pro plan, agents also test iOS and Android builds on cloud iPhones and Android phones through Revyl. Capy does not describe approvals for purchases or form submissions made in the browser. Sourcedocs.capy.ai/machinesread 2026-10-28 |
||
Pricing snapshot
Sourced from the Index pricing dataset · open each vendor's profile for full detail.
| Pricing | G Google Jules |
C Capy |
|---|---|---|
|
Entry price Lowest public entry point |
Free; Jules in Pro with Google AI Pro at $19.99 a month | Free on your own computers; Lite $20 a month, or $16 billed annually; Pro $100; Max $200; Enterprise custom |
|
Pricing confidence How public the numbers are |
Public, exact | Public, exact |
|
Billing Primary billing axis |
Daily task quotas set by a Google AI subscription | One subscription per organization with a shared monthly credit balance and no seat fee. Model tokens and cloud machine time draw the credits down. |
|
Variable cost Workload / overage exposure |
Low variable cost | High variable cost |
|
Free tier / trial Try before you buy |
Free tier
|
No free tierTrial
|
|
Buying motion Self-serve vs sales call |
Self-serve | Mixed |
|
Pricing verified When this pricing was last checked |
Product updates
Newest first, each linked to its source.
Google Jules
-
· Agent capability
Sessions in Jules can continue until November 11, 2026, after which Jules turns read only, and the product winds down for good on December 31, 2026 as Google unifies its agentic coding work.
Source
Capy
-
· Security / enterprise
Version 0.8.0 of Capy adds an Identity and access page in organization settings where admins verify a domain by DNS record and connect Google Workspace sign in, sold as an add on, with Microsoft Entra ID, Okta and other SAML or OIDC providers by contact.
Source -
· Deployment / data residency
The new Dock CLI in Capy 0.7.0 runs Capy without the desktop app, so a build server, cloud VM or spare Mac can host agent threads with the tools and credentials already on it.
Source -
· Workflow orchestration
Capy 0.4.7 lets a Captain, the agent that coordinates a crew of other threads, run on one of the user's paired devices, and the Captain can start crew threads on any paired device or in the cloud.
Source
Jules is a Google Labs product, and Google announced its wind down on 29 October 2026. Capy is built by Scrapybara, Inc., a Y Combinator company that earlier sold hosted computers for agents, and its first public release came in August 2026. Jules never offered organization administration, single sign on or a compliance attestation, while Capy holds SOC 2 Type II. Capy's hosted machines run in one region, US East. Neither prints what a typical task costs, so ask Capy how model usage and machine minutes would cover the work Jules used to do.
More comparisons with Google Jules or Capy
Other matchups in coding agents
Not the pairing you were after? These compare a different set of coding agents on the same 14 capabilities.