Arcreel
AI Agent 驱动的开源视频生成工作台 — 小说→角色/场景/道具设计→剧本→分镜图→视频，跨镜头角色与场景一致 | Open-source AI video workspace powered by AI Agents, Nano Banana 2 & Veo 3.1 / Grok / Seedance / OpenAI
Last verified:
What is Arcreel?
ArcReel is an open-source AI short-drama and manga workspace that transforms novels into finished video reels through multi-agent orchestration. Users drop in a novel (Chinese or English, any length), and AI agents automatically handle scriptwriting, character design, storyboarding, and final video synthesis while maintaining visual consistency for characters, scenes, and props across every shot.
The platform features a state-aware orchestrator that dispatches focused subagents for each task, using the Claude Agent SDK's Skill + Subagent pattern. Key capabilities include cross-shot character consistency through reference sheets, an async task engine with RPM-rate-limited scheduling, one-click version rollback, and export to Jianying desktop for human finishing. The pipeline has 6 stages: upload novel, build asset library, plan & split episodes, generate script JSON, create character & storyboard frames, and synthesize video.
ArcReel supports 5 preset providers (Gemini, Volcengine, Grok, OpenAI, Vidu) with 40+ models across image, video, and text modalities, plus unlimited custom OpenAI-compatible or Google-compatible endpoints including self-hosted Ollama and vLLM. It is built for content creators, video production companies, developers, and anyone wanting to automate video production from text. The software runs on Docker via Linux, macOS, or WSL2, and is released under AGPL-3.0 license.
The tool is particularly suitable for creating promotional videos, educational content, AI manga dramas (anime/animation style), AI short dramas (live-action realism), and narration videos. It addresses pain points in AI video creation such as poor cross-shot consistency, uncontrollable costs, and fragmented workflows through its comprehensive automation approach.
Arcreel pricing
Pricing model: Freemium
The open-source build is AGPL-3.0 and free forever - self-host with Docker Compose on Linux, macOS, or WSL2. Users only pay for upstream AI provider usage (Gemini, Volcengine, Grok, OpenAI, or their own custom endpoints). The commercial tier is an optional managed offering that adds hosted infrastructure, pooled provider credits at better rates, priority GPU queues, team workspaces with SSO, white-label branding for agency use, and a support SLA. Limited spots are available in the current commercial beta cohort.
Arcreel pros
- Open-source and free forever for self-hosted use
- Maintains cross-shot character, scene, and prop consistency
- Supports 5 providers with 40+ models out of the box
- Allows unlimited custom endpoints including self-hosted Ollama and vLLM
- Multi-agent orchestration with Claude Agent SDK architecture
- One-click version rollback with side-by-side variant comparison
- Async task engine with resumable downloads and RPM rate limiting
- Exports Jianying-ready drafts for desktop finishing
- Supports both Chinese and English novels of any length
- Two content modes: narration and drama
- Progressive human-in-the-loop episode breakdown with AI suggestions
- No GPU required when using preset cloud providers
- Cost tracking built into the pipeline
- AGPL-3.0 license allows commercial use
- Data stays on your machine - full privacy control
- Model-neutral - switch between Sora 2, Veo 3.1, Seedance 2.0 freely
- Docker-based deployment runs on Linux, macOS, WSL2
Arcreel cons
- No native Windows support - requires WSL2 or Docker Desktop
- You pay separately for upstream AI provider usage
- Requires POSIX-like environment (Linux/macOS/WSL2)
- Commercial beta has limited spots and is optional
- AGPL-3.0 copyleft requires releasing modifications if served over network
- No zero-ops ready templates like closed SaaS platforms
- Self-hosting requires Docker Compose setup knowledge
- Local LLM support needs GPU if using Ollama/vLLM
- Commercial beta still accepting applications - not fully open yet
Frequently asked questions about Arcreel
Is ArcReel free to use?
Yes, the open-source build is AGPL-3.0 and free forever. You self-host with Docker Compose on Linux, macOS, or WSL2. You only pay for the upstream AI provider usage (Gemini, Volcengine, Grok, OpenAI, or your own custom endpoints). The commercial tier is optional for teams wanting hosted infrastructure, pooled credits, and SLAs.
How does ArcReel maintain character consistency across shots?
Before any storyboard is rendered, a subagent scans the novel and builds a library of characters, scenes, and props. Reference sheets are generated first, and every downstream image/video generation is conditioned on them. Characters, scenes, and props are tracked as persistent assets across every cut. Continuity is designed into the pipeline, not a prompt hack.
Which AI providers and models does ArcReel support?
Five preset providers ship out of the box: Gemini (Nano Banana 2 / Veo 3.1), Volcengine (Seedream / Seedance), Grok, OpenAI (GPT Image / Sora 2), and Vidu (Vidu Q3 video / Vidu Q2 image) - totaling 40+ preset models across image, video, and text. You can also add any OpenAI-compatible or Google-compatible endpoint, including self-hosted Ollama and vLLM, and ArcReel will auto-discover models.
Can I export to video editing software?
Yes. Per-episode drafts export as a ZIP compatible with Jianying desktop 5.x and 6+. You import the ZIP, fine-tune the cut, add music, and you're done. There's also an FFmpeg pipeline for automated cuts.
What operating systems does ArcReel run on?
ArcReel requires a POSIX-like environment: Linux, macOS, or Windows with WSL2/Docker Desktop. A few low-level dependencies are POSIX-only, so native Windows isn't supported yet.
What's the difference between open-source and commercial versions?
Same core pipeline, but the commercial version is hosted - no setup required. It adds pooled provider credits at better rates, priority GPU queues, team workspaces with SSO, white-label branding for agency use, and a support SLA. The open-source version requires self-hosting but is free forever.
How does ArcReel compare to closed-source platforms like Runway or Pika?
Closed-source SaaS platforms keep your scripts, character assets, and generated content in their cloud. ArcReel is AGPL-3.0 open source and self-hostable: data stays on your machine, you hold the model API keys, and full novels never leave your server. ArcReel is also model-neutral with Sora 2, Veo 3.1, Seedance 2.0, Nano Banana 2, GPT Image 2, and Vidu Q3 wired up out of the box. Commercial platforms win on zero-ops and ready templates; ArcReel wins on control, hackability, and zero subscription fees.
Which video model should I use: Veo 3.1, Sora 2, or Seedance 2.0?
All three ship preset in ArcReel. Pick by use case: Veo 3.1 produces the most natural long takes and camera motion, with Fast/Lite variants for speed-cost tradeoffs. Sora 2/Sora 2 Pro give the strongest realism and acting detail. Seedance 2.0 has the richest multi-modal reference (mix up to 9 images, 3 videos, 3 audio clips per shot), strong Chinese-scene understanding, and token-based pricing that stays predictable on long projects. Vidu Q3 is a fourth option with reference-to-video plus audio generation. Switch globally or per-project - no need to commit upfront.
Do I need a GPU to run ArcReel?
No. All image/video/text generation goes to provider APIs (Gemini, Seedance, Sora, Grok, etc.) in the cloud. Your machine only runs orchestration logic, FFmpeg compositing, and Jianying draft packaging. A normal laptop with macOS/Linux/WSL2 + Docker Compose is enough. If you wire in self-hosted Ollama/vLLM for local LLMs, that machine needs a GPU, but with the five preset providers, no GPU is required.
Can I use custom AI models or providers?
Yes. ArcReel uses unified ImageBackend/VideoBackend/TextBackend protocols - anything OpenAI-compatible or Google-compatible plugs in. Add a custom provider in settings with Base URL + API Key; ArcReel hits /v1/models to discover what's available and infers media type by name. Custom providers get the same treatment as the five presets: project-level switching, cost tracking, version history. Third-party API gateways work too.