Anthropic Claude Code vs JetBrains AI (2026)

Upgraded from P: terminal, desktop, IDE, web, GitHub Actions, Remote Control from mobile, and a headless -p flag for scripting are seven distinct invocation paths, plus hooks firing on lifecycle events.

Upgraded from P: invocation now spans IDE chat, tool window, standalone CLI in any terminal, CI/CD via GitHub Actions and GitLab, ACP clients, and remote or async runs, which is breadth rather than a single entry point.

Upgraded from P: grounding runs on the IDE's semantic analysis engine rather than text embeddings, giving import graphs, class hierarchies and call chains, plus persistent AGENTS.md instructions and MCP connected external sources.

Upgraded from P: CLAUDE.md is durable repository scoped memory read every session, session transcripts persist and are resumable, and Remote Control syncs a transcript across devices.

AGENTS.md is durable project scoped instruction context rather than agent memory; sessions are managed and resumable but no cross session memory store is documented.

Held at P: .aiignore, AGENTS.md and per command approval are real user level controls, but no organisation level identity or governance surface (SSO, RBAC, audit logging, admin policy) was retrieved for the AI product, and no security attestation specific to it. The axis conjunction is not met on the evidence found.

Upgraded from P: OpenTelemetry export with tool level detail is an organisation grade audit trail, and cloud sessions log all operations for compliance. Hooks make the trail programmable, which is stronger than the reporting-only surfaces I downgraded elsewhere in this lane.

Progress reporting, reviewable diffs and session status are documented, but no persistent execution trace or audit log across sessions; the debugger session state is inspectable at runtime rather than retained as a record.

Upgraded from P: self hosted environments for cloud sessions and customer chosen inference platforms are both documented, alongside the default of running entirely on the developer's own machine. Recorded honestly: the docs state data is encrypted in transit but not at rest, and Zero Data Retention is an enterprise arrangement rather than a default.

Local model serving via Ollama, LiteLLM or LMStudio keeps inference on the developer's own machine, which is real deployment control at the individual level; held at P because no organisation level self host, VPC or residency option was retrieved.

Upgraded from N: plugins are a versioned installable unit distributed through a public marketplace with a discovery and browse UI, plus bundled skills shipped by Anthropic. That is the axis, and it did not exist in this form when the June record was built.

Modes and slash commands are behaviour presets rather than a library of prebuilt agents or installable packs; the MCP registry is a server catalogue, credited to Ext.

Upgraded from N. The axis measures whether the customer chooses the model powering the product. Bedrock, Vertex and Foundry are customer selected inference platforms under the customer's own cloud account and credentials, and model selection plus fallback-model routing are documented, with hooks able to block or annotate a model switch. All models are Anthropic's, which is a real constraint and is recorded here, but provider and deployment plane choice plus per subagent model selection clears the F bar as applied to activepieces and coderabbit in this lane.

Downgraded from F, applying the same axis rule used on coderabbit, kiro, google-antigravity and jetbrains-ai in this lane. /verify, /run and /code-review test the customer's code and confirm the agent's change works, which is self verification of output; no harness for evaluating or regression testing agent behaviour itself is documented. Graded consistently rather than favourably.

Debug mode is the strongest evidence on this axis in the lane so far: the agent inspects live runtime state through the real debugger rather than reasoning from logs, which is genuine verification of its own work.

Meets the axis test squarely: the agent drives real web applications through their human interface, using the customer's existing logged in session, reaching software that exposes no programmatic interface to it. Confirmed on Anthropic's own docs rather than the unverifiable sub-agent browser surface reference the June basis cited.