Everything you make, on one canvas.
Make pictures, video, 3D, sites, slides and posts with the best models, alongside your team and the agents you already use.
Fil is invite-only for now. Join the waitlist, or log in if you have an invitation.
Every leading model, one credit balance
58 models for pictures, video, 3D, pages and sound. Each says what it costs before it runs.
Images
- GPT Image 2.5 Flare by OpenAI: fast, four variants, transparent PNGs
- GPT Image 2.5 Sunburst by OpenAI: precise edits
- Ideogram 4.5 by Ideogram: type, layout and edits
- Nano Banana 2.1 by Google: plans before it draws
- MAI-Image-2.6 by Microsoft AI: text, portraits, precise edits
- MAI-Image-2.6 Flash by Microsoft AI: fast, low cost
- Seedream 5.0 Pro by ByteDance
- Seedream 5.0 Flash by ByteDance: fast, low cost
- FLUX 3 Image by Black Forest Labs: edits, up to 4K
- Muse Image by Meta: edits, looks things up
- Grok Imagine 2.0 by xAI
- Soul 2 by Higgsfield
- Qwen Image 3 by Alibaba: text in English and Chinese, edits
Video
- Grok Imagine Video 1.5 by xAI: with audio
- Grok Imagine Video 1.5 Lite by xAI: faster, lower cost
- FLUX 3 Video by Black Forest Labs: with sound, up to 4K
- Seedance 2.0 by ByteDance
- Gemini Omni Flash by Google: with sound, 3–10 s
- Seedance 2.5 by ByteDance: up to 30 s, with audio
- Seedance 2.0 Mini by ByteDance: fast, low cost, with audio
- Seedance 2.0 Fast by ByteDance: faster, with audio
- Kling 3.0 by Kling: with sound, up to 4K
- Kling 4.0 by Kling: up to 30 s, with sound
- MiniMax H3 by MiniMax: Hailuo 03, 2K
- MiniMax H3 Max by MiniMax: faster, 768p
- Wan 3.0 by Alibaba: up to 30 s, with audio
- Wan 3.0 Prime by Alibaba: faster, up to 30 s, with audio
- HappyHorse 1.1 by Alibaba: up to 1080p and 15 s
Pages and code
- Claude Opus 5.5 by Anthropic
- Claude Fable 5.1 by Anthropic: most capable
- Claude Sonnet 5.5 by Anthropic: faster
- Claude Haiku 5.5 by Anthropic: fastest, cheapest
- GPT-6 Astra by OpenAI
- GPT-6.1 Sol by OpenAI: faster
- Gemini 4 Argon by Google: most capable
- Gemini 3.8 Flash by Google: fast, low cost
- Muse Wonder 1.0 by Meta: most capable
- Muse Spark 1.3 by Meta: low cost
- Grok 4.7 by xAI
- Kimi K3.1 Pro by Moonshot AI: most capable
- Kimi K3.1 Pro Highspeed by Moonshot AI: faster
- Kimi K3 by Moonshot AI
- MiniMax M3 by MiniMax: low cost
- MiniMax M3.1 Flash by MiniMax: fast, low cost
3D
- Tripo 3.1 by Tripo: textured GLB
- Tripo 2.5 by Tripo: balanced
- Meshy 7.1 by Meshy: textured GLB, PBR
- Meshy 6 Lite by Meshy: faster, lower cost
Voice and sound
- ElevenLabs v3 by ElevenLabs: expressive voiceover
- ElevenLabs v4 by ElevenLabs: most expressive voiceover
- ElevenLabs v4 Turbo by ElevenLabs: fast voiceover, lower cost
- ElevenLabs Multilingual v2 by ElevenLabs: your cloned voice, 29 languages
- GPT-4o mini TTS by OpenAI: voiceover
- Gemini 3.8 Flash TTS by Google: voiceover
- Grok TTS by xAI: expressive voiceover
- ElevenLabs Sound Effects by ElevenLabs: 0.5 to 30 s, seamless loops
- ElevenLabs Music by ElevenLabs: songs up to 5 min
- Lyria 3.5 by Google: full songs
What Fil does
- Images: Make images up to 4K with 13 AI models from OpenAI, Ideogram, Google, Microsoft AI and more, then paint over, extend or upscale them on a canvas.
- Videos: Make video clips with sound from a sentence or a still with 15 AI models from xAI, Black Forest Labs, ByteDance, Google and more, then edit them into one video.
- 3D: Make textured 3D models (GLB) from a prompt or a picture with Tripo and Meshy, then rig, animate and edit them in Fil’s 3D editor.
- Agents: Connect Claude Code, Codex, Cursor or ChatGPT to Fil’s MCP server: your agent makes images, video, 3D and sites on your canvases, with your credits.
- Pricing: Free with 100 credits a day. Starter is $18 a month, Pro $30 and Max $100 a seat. Every run shows its price before it starts, and failed runs are refunded.
- Templates: Start from a canvas that’s already set up: YouTube covers, UGC videos, portraits, websites, social posts, 3D models, brand identities and logos.
41 workflows
Pick a workflow, say what you want, and Generate all makes every step in turn, from the brief to the finished piece.
- Logo and brand guide
- Manga chapter
- Story video
- Product promo video
- Long-form product video
- Product launch video
- Motion film
- Product photoshoot
- 3D asset
- Game character
- Character gear
- Material maps to Three.js
Community
What people make with Fil, with the prompt and model behind each piece.
Plans
Every plan has every model and resolution, and people and agents spend the same credits.
Free: $0
For trying it out.
- 100 credits a day when you check in
- 200 on every third day in a row
- Up to 1,000 credits a month
- Daily credits last 24 hours
- 1 GB of storage, 1 GB of bandwidth a month
- Uploads up to 1 MB a file, sites up to 2 MB
- 3 canvases and 1 project
- Every model and resolution
- Agents over MCP
- What you make is public on Explore
Starter: $18 / month
For one person who generates every day.
- 1,800 credits a month
- 20 GB of storage, 50 GB of bandwidth a month
- Uploads up to 10 MB a file, sites up to 20 MB
- Sites at an address you choose
- Unlimited canvases and projects
- Private runs
- Every model and resolution
- Agents over MCP
- Commercial licence
Pro: $30 / seat / month
For teams that make things together.
- 3,000 credits a month for each seat, in one pool
- 50 GB of storage, 100 GB of bandwidth a month for each seat
- Uploads up to 50 MB a file, sites up to 50 MB
- A team with a shared library and canvases
- Editor and viewer permissions
- Sites at an address you choose
- Private runs
- Commercial licence
Max: $100 / seat / month
For studios that generate all day.
- 10,000 credits a month for each seat, in one pool
- 200 GB of storage, 250 GB of bandwidth a month for each seat
- Uploads up to 100 MB a file, sites up to 100 MB
- Everything in Pro