Free AI for Open Source
A grant of 15,000,000 free AI tokens plus Pro-tier API access for maintainers of actively maintained open source projects.
100% Open-Source AI Models
Ogo-onwe-n'onwe ya na-echekwa móòdù niile na Free.ai bụ ófèèsì n'okpuru ikikembanye (Apache 2.0, MIT, BSD). Enweghị ikikembanye Llama/Meta. Anyị na-eche na AI kwesịrị ịbụ nke na-egosi, nke a ga-enyocha, nakwa nke ọbụla iji.
Chat Ụdị asụsụ
| Móòdù | Jiri ihenhọrọ ahụ | Ikikere | Òtù |
|---|---|---|---|
| Qwen2.5-72B | Ndụmọdụ, nghọta, nlekọta | Apache 2.0 | HuggingFace |
| Qwen2.5-7B | Nkata n'ụzọ nkịtị, ọrụ dị n'okpuru | Apache 2.0 | HuggingFace |
| Mistral-7B | Nkata n'ụzọ nkịtị | Apache 2.0 | HuggingFace |
| Phi-3 | Nkọwapụta dị mma, mathematics | MIT | HuggingFace |
Kóòdù Models
| Móòdù | Jiri ihenhọrọ ahụ | Ikikere | Òtù |
|---|---|---|---|
| Qwen2.5-Coder-32B | Nhazi koodị, debugging | Apache 2.0 | HuggingFace |
| DeepSeek-Coder-V2-Lite | Nkọwapụta koodị, nlele | MIT | HuggingFace |
Inyogo na vidio
| Móòdù | Jiri ihenhọrọ ahụ | Ikikere | Òtù |
|---|---|---|---|
| FLUX.1-n'ihi | Nhazi inyogo n'ụzọ nkịtị | Apache 2.0 | HuggingFace |
| Stable Diffusion XL | Ihu ndị ahụ dị elu | OpenRAIL++ | HuggingFace |
| Kandinsky | Ihu ndịna | Apache 2.0 | HuggingFace |
| CogVideoX-2B | Vidéòdù nke a kpụgara site na ngwe | Apache 2.0 | HuggingFace |
Nsụgharị
| Móòdù | Jiri ihenhọrọ ahụ | Ikikere | Òtù |
|---|---|---|---|
| Kokoro | Nsọ | Apache 2.0 | HuggingFace |
| Piper | Nsọ | MIT | GitHub |
| MeloTTS | Nsọ | MIT | GitHub |
| Chatterbox | Nsọ | MIT | HuggingFace |
| Òtù | Nsọ | Apache 2.0 | HuggingFace |
| faster-whisper | Nsụgharị | MIT | GitHub |
| Òsì | Nsụgharị | Apache 2.0 | GitHub |
| ACE-Step 3.5B | Nhazi egwu | Apache 2.0 | HuggingFace |
| Mel-Band RoFormer | Nhazi egwu (n'asụsụ ọbụla, n'ime ihenhọrọ) | MIT | arXiv |
| SCNet | Nhazi egwu (n'ime drums, bass, ndị ọzọ) | Apache 2.0 | arXiv |
Ntụgharị, Nlegharị anya na Móòdù ndị ọzọ
| Móòdù | Jiri ihenhọrọ ahụ | Ikikere | Òtù |
|---|---|---|---|
| MadLAD-400 3B | Ntụgharị (450+ asụsụ) | Apache 2.0 | HuggingFace |
| Real-ESRGAN | Nhazi inyogo | BSD | GitHub |
| BRIA RMBG 2.0 | Wepụ okpuru | Apache 2.0 | HuggingFace |
| US | Nhazi inyogo | Apache 2.0 | GitHub |
| De_oldify | Nhazi ụcha foto | MIT | GitHub |
| all-MiniLM-L6-v2 | Ńkwádò ngwe | Apache 2.0 | HuggingFace |
| Tesseract | OCR | Apache 2.0 | GitHub |
Open Source
Ndị na-ahụ maka ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ nwere ike ịnyefe maka nkwekọrịta n'efu nke 15,000,000 tokens, nkwekọrịta nkwekọrịta nke afọ nke Pro, yana API keys na Pro request tier (100,000 nkwekọrịta kwa ọnwa). The grant covers every tool that runs on our self-hosted models. Premium external models are not part of the grant: they use purchased tokens.
Onye na-ekwenyere
- Í na-echekwa ákàrà ahụ (onye nwe ya mọọbụ onye na-eweta ya), ná repository ya bụ nke ọha na eze n'okpuru OSI-ekwenyere ná ikikembanye ohuru ohuru
- Ọ na-echekwa ya n'ụzọ dị irè, na-eji n'ime ụbọchị 90 gara aga
- Ndị ọzọ na-eji ya, egosipụtara site n'otu n'ime: 50 GitHub stars, 1,000 package downloads a month, or projects that depend on it
- Otu n'ime ego ndị ahụ n'ime akaụntụ Free.ai n'ime ebe obibi, n'enweghị uru nakwa usoroiheomume mbido, agbakwunyere na akaụntụ ịkewapụtara
Iji wụnye, zipụ njikọ repository na njikọ na-egosi na ịrụzi (stars, download statistics or dependents) site na adreesị ozi-e na akaụntụ gị Free.ai. Wepụta site n'aka onyeọbịa
Ajụjụ ndị a na-ajụkarị
Yes. Every model we run on our GPU servers is open-source under permissive licenses -- Apache 2.0, MIT, or BSD. We deliberately exclude models with restrictive licenses like Meta's Llama license. You can inspect, use, and modify all models freely.
Meta's Llama license includes restrictions on commercial use for large companies and requires specific attribution. We only use models with truly permissive licenses (Apache 2.0, MIT, BSD) that allow unrestricted commercial and personal use by anyone.
Every model in our table links to its source on HuggingFace or GitHub. You can download the weights directly from these repositories and run them on your own hardware.
Qwen2.5-72B is our most capable chat model for complex reasoning and analysis. Qwen2.5-7B and Mistral-7B are faster options for everyday tasks. Phi-3 is excellent for efficient reasoning and math.
We use five text-to-speech models: Kokoro (Apache 2.0), Piper (MIT), MeloTTS (MIT), Chatterbox (MIT), and OuteTTS (Apache 2.0). Each has different strengths in voice quality, speed, and language support.
FLUX.1-schnell for fast generation (Apache 2.0), Stable Diffusion XL for high-quality images (OpenRAIL++), and Kandinsky 2.2 for artistic styles (Apache 2.0). For video, we use CogVideoX-2B, Wan 2.2 and HunyuanVideo.
If you have developed or know of an open-source model under a permissive license that you think we should host, contact us. We evaluate models based on quality, license compatibility, and user demand.
A free grant nke 15,000,000 tokens plus API keys na Pro request tier (100,000 requests per month) maka ndị na-echekwa nke na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ na-
We evaluate every model before deploying it based on benchmark performance, real-world output quality, inference speed, and license compliance. Only models that meet our quality bar are added to the platform.
We use MadLAD-400 3B from Google, which supports 450+ languages under an Apache 2.0 license. It handles even low-resource languages that many commercial translation services do not support.
Real-ESRGAN (BSD) for image upscaling, BRIA RMBG 2.0 (Apache 2.0) for background removal, SAM (Apache 2.0) for image segmentation, and DeOldify (MIT) for photo colorization.
Ọ dịghị: premium external models dị ka GPT-5, Claude na Gemini bụ proprietary, na-eburu iwu ikike ha, na-agba ọsọ na tokens na-ere ahịa na ọ bụghị akụkụ nke ọ bụla comp grant. Only the self-hosted models listed on this page are open source.