Flowith
Flowith is an agentic AI workspace that combines a whiteboard-style canvas, access to 40+ image and video models, and mini apps for design, e-commerce, and social content.

Junie is JetBrains' AI coding agent for developers who want model choice, structured planning, and real-time control over autonomous coding tasks.
Junie is JetBrains' AI coding agent for developers who want to choose their own models, plan before code is written, and stay in control while the agent works. It runs in the terminal and inside JetBrains IDEs, and it is positioned as a cost-efficient alternative to agents that lock you into a single model or a fixed subscription.
Junie's defining idea is that planning and implementation do not have to run on the same model. You can plan on a powerful model and implement on a faster one, which JetBrains says delivers the same quality at a fraction of the cost. The agent supports more than 10 models through Bring Your Own Key, including Claude Opus 5.5, Claude Sonnet 5.5, Claude Fable 5.1, Gemini 3.1 Pro, Gemini 3.8 Flash, GPT-6 ASTRA, GPT-6.1 SOL, and Grok 4.7. It also supports locally running models. Because BYOK uses provider-rate pricing with zero markup, the cost structure is transparent, though AI usage is billed separately from JetBrains AI subscriptions.
Before Junie touches code, Advanced Plan Mode writes structured plan requirements, design, and delivery stages. Those plans live in .junie/plans, where they are editable and committable, so they can be reviewed like any other project artifact. You can approve a plan or redirect it. The separation between planning and implementation is what makes the two-model strategy practical: the expensive reasoning happens once, and the execution can run on a cheaper model.
Junie is built around the assumption that developers want to steer, not just watch. Live Prompting lets you change direction mid-task without starting over, a feature JetBrains describes as exclusive. Human-in-the-loop controls add a dynamic allowlist and require user confirmation before key actions execute. Together, these features reduce wasted tokens and keep the agent from running too far in the wrong direction.
Junie can be taught your team's coding standards, naming conventions, and review rules. Those guidelines persist across sessions, and Agent Skills packages extend the agent's capabilities. Slash commands are shared across the CLI and IDE through ACP, so the same instructions work wherever you run Junie. This matters for teams that want consistent output without repeating the same context in every session.
Junie Lite is a free model available in every JetBrains account, with no trial and no credit card required. For heavier use, JetBrains AI Pro costs $8.33 per user per month and includes 10 AI credits per 30 days, while AI Ultimate costs $25.00 per user per month and includes 35 AI credits per 30 days. AI Ultimate is recommended for regular Junie use. Anytime top-ups are available, and AI usage is billed separately. Prices and credit allowances can change, so verify current terms on the official JetBrains pricing page.
JetBrains states that its tools adhere to industry-leading security standards, including SOC 2 certification. For teams switching from another AI coding agent, JetBrains advertises hassle-free migration with minimal setup and no workflow disruption. That claim is worth testing against your own repository and CI setup before committing to a full switch.
Junie is most compelling for developers who already use JetBrains IDEs and want an agent that respects model choice and human review. The combination of Advanced Plan Mode, Live Prompting, and persistent guidelines addresses common complaints about autonomous coding agents: opaque plans, runaway execution, and repeated context. The trade-off is that the best results depend on the models you bring, and those models carry their own costs. Junie Lite lowers the barrier to entry, but teams that rely on it heavily will likely need a paid AI plan or BYOK arrangement.
The first thing to understand is that Junie has no price tag of its own. You never buy "Junie premium" as a separate product β Junie is the coding agent inside the JetBrains AI subscription, and everything it does is paid for out of a shared pool of AI Credits, the same pool used by AI chat in your IDE. One credit corresponds to roughly one US dollar of model usage, and agentic tasks β where Junie plans, edits files and runs tests over many steps β naturally spend credits faster than a quick chat question.
There are three ways in. Junie Lite is the free starting point: a free model tier available in every JetBrains account, with no trial period and no card required, and you can also bring your own API key (BYOK) and pay your model provider directly at provider rates with zero markup. AI Pro costs $8.33 per user per month on annual billing ($10 monthly) and includes 10 AI Credits per 30 days β enough to try Junie properly and use it for lighter tasks. AI Ultimate, at $25 per user per month billed annually ($30 monthly), raises that to 35 credits per 30 days and is the tier JetBrains itself recommends for regular Junie users. Both paid tiers allow anytime top-ups if you run dry mid-month.
Which tier you need depends less on headcount than on how autonomous you let Junie be. A developer who asks Junie to implement a planned feature end-to-end a few times a week will feel the 10-credit Pro allowance quickly; teams running the agent daily across large codebases should budget for Ultimate or top-ups from the start. The sensible path is to begin with Junie Lite or BYOK β where your only cost is the model bill itself β measure how much agent work you actually use, and then let that number pick the tier.
Is JetBrains Junie free? Partly. Junie Lite is a free model tier included in every JetBrains account β no trial and no card required β and with bring-your-own-key (BYOK) you can run Junie on your own OpenAI or Anthropic API key. Sustained use of JetBrains' cloud models runs on AI Credits from a paid AI Pro or AI Ultimate subscription.
How much does Junie cost? Junie is not sold separately; it draws on the JetBrains AI subscription. AI Pro is $8.33 per user per month billed annually ($10 monthly) with 10 AI Credits per 30 days. AI Ultimate is $25 billed annually ($30 monthly) with 35 credits per 30 days, and both tiers support paid top-ups.
Which IDEs does Junie work in? Junie runs inside the main JetBrains IDEs β IntelliJ IDEA, PyCharm, WebStorm, GoLand, PhpStorm, RubyMine, RustRover, Rider and CLion among them β and also as the standalone Junie CLI in a terminal or CI/CD pipeline.
What is the difference between Junie and JetBrains AI Assistant? AI Assistant is the broader AI layer in JetBrains IDEs β chat, completion and explanations. Junie is the autonomous agent within that ecosystem: you hand it a task, it writes a structured plan, edits code across files, runs tests and asks for your approval at key steps before continuing.
Junie Lite is free in every JetBrains account with no trial or card required. Bring Your Own Key offers provider-rate pricing with zero markup. JetBrains AI Pro costs $8.33 per user per month and includes 10 AI credits per 30 days. AI Ultimate costs $25.00 per user per month and includes 35 AI credits per 30 days. AI usage is billed separately. Prices and credit allowances are subject to change.
Features, pricing and availability may change. Always verify details on the official website. Maintained by steven Β· Last updated Nov 10, 2026 β
Flowith is an agentic AI workspace that combines a whiteboard-style canvas, access to 40+ image and video models, and mini apps for design, e-commerce, and social content.
Fine-tune open models faster with far less memory
Groq is an AI inference cloud for developers and enterprises that need low-latency model responses at scale, delivered through custom LPU hardware and a tiered platform spanning bare metal, inference, and enterprise controls.
Jan is an open-source desktop chat application that lets users run open models on their own hardware or connect to hosted providers like OpenAI, Anthropic, and Google. It targets privacy-conscious individuals and teams who want a friendlier local LLM experience than command-line tooling offers.