Standout features
Wan is Alibaba’s open-source text- and image-to-video model — Apache-2.0 licensed, free to self-host or run via API, and consistently top of open-source video benchmarks.
Worldwide search interest, indexed 0–100 · Google Trends.
Wan is Alibaba’s open-source video model — free for commercial use, self-hostable, and the strongest open option, closing the gap with closed models.
- By Alibaba’s Tongyi Lab; open-sourced under Apache 2.0.
- Wan 2.1 (2025) → 2.2 → newer suites; millions of downloads.
- Tops public open-source video benchmarks.
- Self-host on your own GPUs, or run via API.
Wan is an open model family.
- Text-to-video and image-to-video.
- Native 1080p on newer versions; MoE architecture.
- Apache 2.0 — free for commercial use.
- Big community on Hugging Face / ModelScope.
Free weights; you pay for compute.
Wan fits builders + tinkerers.
- Developers wanting open weights.
- Teams avoiding per-clip licensing.
- Researchers + self-hosters with GPUs.
- Non-technical users wanting a polished app.
- Anyone needing one-click cloud simplicity.
No tool is perfect — the trade-offs to weigh:
- Setup + GPU needed to self-host.
- No polished consumer app from Alibaba.
- Audio + multi-shot trail the best closed models.
- Quality varies by version + settings.
- ✓Open weights (Apache 2.0)
- ✓Free for commercial use
- ✓Tops open-source benchmarks
- ✓Self-host or via API
- ✓Large active community
- ✕Needs GPU + setup
- ✕No polished consumer app
- ✕Audio / multi-shot trail closed models
- ✕Quality varies by version
Builders love Wan for being genuinely open — Apache-2.0 weights, free for commercial use, top of open-source benchmarks, and self-hostable with no per-clip fees. The gripes are the GPU/setup barrier, no polished app, and audio/multi-shot lagging the best closed models. Sentiment is very positive among developers.
Wan is the open-source video model family from Alibaba’s Tongyi Lab.
Company figures are drawn from public disclosures and reputable trackers (gathered Jun 2026). User and revenue numbers are estimates and move fast.
Pick up to two other coding tools to see them head-to-head on the same rubric.