JCode

JCode is an open-source coding agent designed to work with local development workflows and optional hosted-model inference. The project has no feature gates in its open-source agent; its hosted model service uses an inference-credit and subscription model. Developers can use the agent to inspect repositories, plan changes, edit files, run development tasks, and choose how inference is supplied.

✨ Key Features

• Use an open-source coding agent — Run the full agent without feature gating.
• Work in a repository — Inspect project context and reason about code changes.
• Plan implementation tasks — Break a coding request into actionable steps.
• Edit code with agent assistance — Apply changes across a codebase through an agent workflow.
• Run development actions — Use the agent to carry out supported commands and checks.
• Choose inference providers — Use local/provider API options or the hosted model service.
• Use hosted-model credits — Apply inference credits to hosted usage.
• Keep provider control — Bring your own model/provider when the workflow supports it.
• Inspect open-source code — Review and contribute to the project under its open license.
• Use in existing engineering workflows — Keep the agent close to normal development tools and repositories.

💳 Pricing

• Open-source agent — Free under the MIT license; use your own API keys, supported OAuth subscriptions, or local models. Provider costs remain separate.
• Hosted subscription — Choose a monthly plan from $10 to $1,000. Each payment grants inference credits worth twice the subscription amount and 50 Cloud machine hours per $10 of plan.
• Hosted inference — Included credits are used first; subsequent usage is billed at the provider's API rates, in addition to the subscription. The automatic additional monthly usage cap is 10 times the plan amount: $100 for a $10 plan. Account settings can change the cap.

Hosted inference and Cloud access are optional; confirm provider, machine-hour, usage-cap, and billing terms before subscribing.

⚖️ Alternatives Compared

🎯 Use Cases

• Exploring an unfamiliar repository.
• Planning a multi-file implementation.
• Applying repetitive refactors with agent assistance.
• Running development checks after an edit.
• Using an open-source agent in a self-managed workflow.
• Choosing a preferred inference provider.
• Testing hosted inference with a credit allowance.
• Reviewing and contributing to an agent project.