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-Bron-KImodel's
Elke eie-geëneerde model op Free.ai is oop-onsece onder permissief lisensies (Apache 2.0, MIT, BSD). Geen Llama/Meta lisensies. Ons glo dat Kunsmatige intelligensie deursigtig, ouditeurent en vry is om te gebruik.
Chat & Language Modelle
| Model | Gebruik letterkas | Lisensie | Bron |
|---|---|---|---|
| Qwen2.5-72B | Algemene gesels, redeneer, ontleding | Apache 2.0 | HuggingFace |
| Qwen2.5- 7B | Vinnige gesels, ligte take | Apache 2.0 | HuggingFace |
| Mistral- 7B | Vinnige gesels, algemene doel | Apache 2.0 | HuggingFace |
| Phi- 3 | Doeltreffende redenasies, wiskunde | MIT | HuggingFace |
Kommentaarmodel's
| Model | Gebruik letterkas | Lisensie | Bron |
|---|---|---|---|
| Qwen2.5- Coder-32B | Kode geslag, ontfouting | Apache 2.0 | HuggingFace |
| DeepSeek- Coder- V2-Lite | voltooiïng, hersiening | MIT | HuggingFace |
Beeld & video Modelle
| Model | Gebruik letterkas | Lisensie | Bron |
|---|---|---|---|
| FLUX. 1- dol | Vinnige beeld geslag | Apache 2.0 | HuggingFace |
| Stable Diffusion XL | Hoë-kwaliteit beeld geslag | : listbox KDE distribution method | HuggingFace |
| Kandinsky 2,2 | Kunstenaarsbeeldgeslag | Apache 2.0 | HuggingFace |
| CogVideoX-2B | Video-geslag van teks | Apache 2.0 | HuggingFace |
Spraak- en oudiomodel's
| Model | Gebruik letterkas | Lisensie | Bron |
|---|---|---|---|
| Kokoro | Teks- tot- sech | Apache 2.0 | HuggingFace |
| Piper | Teks- tot- sech | MIT | GitHub |
| MeloTTS | Teks- tot- sech | MIT | GitHub |
| Chatterbox | Teks- tot- sech | MIT | HuggingFace |
| EFESE | Teks- tot- sech | Apache 2.0 | HuggingFace |
| faster-whisper | Spraak- tot-teks | MIT | GitHub |
| Vosk | Spraak- tot-teks | Apache 2.0 | GitHub |
| ACE-Step 3.5B | Musiekgeslag | Apache 2.0 | HuggingFace |
| Mel-Band RoFormer | Musiekafskeiing (voorweë, instrumentale) | MIT | arXiv |
| SCNet | Musiekafskeiing (drome, base, ander) | Apache 2.0 | arXiv |
Vertaling, visioen & ander modelle
| Model | Gebruik letterkas | Lisensie | Bron |
|---|---|---|---|
| MadLAD-400 3B | Vertaling (C 450+ tale) | Apache 2.0 | HuggingFace |
| Real-ESRGAN | Beeldopstart | BSD | GitHub |
| BRUA RMBG 2.0 | Agtergrondverwydering | Apache 2.0 | HuggingFace |
| EFESE | Beeld_deling | Apache 2.0 | GitHub |
| Verouderd | Foto kleur | MIT | GitHub |
| al-MiniLM-L6-v2 | Teksomtrekkings | Apache 2.0 | HuggingFace |
| Teserak | Optiese karakter herkenning | Apache 2.0 | GitHub |
Open Bron Onderhouer Program
Onderhouer van aktiewe oop bron projekte kan aansoek doen vir 'n gratis toestaan van 15 000 000 Werkbewyse, die bedrading van' n jaar Pro, plus API sleutels op die Pro versoek-vlak (100 000 versoeke per maand). Die toestaan dek elke instrument wat op ons self-gehostde modelle loop. Premium eksterne modelle is nie deel van die toelae: hulle gebruik gekoopte vormbewyse.
Wie kwalifiseer
- Jy hou die projek (eie of gereelde knoeier) en die stoorplek is publiek onder 'n OSI-apifiseerde oop bronlisensie
- Dit word daadwerklik gehandhaaf, met sy of haar vrystellings in die laaste 90 dae
- Ander mense gebruik dit, wat getoon word deur ten minste een van: 50 GitHubsterre, 1000 pakkie aflaai'n maand af of projekte wat daarvan afhanklik is
- Een toestaan per Free.ai rekening oor die oop bron, sonder om te gebruik en op te stel programme, het bygevoeg by 'n rekening wat jy reeds geskep het
Na wend aan, stuur die stoorplek skakel en 'n skakel na die bewyse van gebruik (sterre, aflaai statistiek of afhanklike) van die e-posadres op jou Free.ai rekening. Pas deur middel van die kontak vorm toe
Constellation name (optional)
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.
'n Vry toestaan van 15 000 000 Werksbewyse plus API sleutels op die Pro versoek-tabel (100 000 versoeke per maand) vir onderhouers van daadwerklik in stand gehoude oop bronne projekte met bewyse van gebruik. Dit dek die gereedskap wat op ons self-gemaatde modelle loop. Premium eksterne modelle is nie deel van die toestaan en gebruik van gekoopde seine. Wend aan met 'n skakel na jou stoorplek en na die bewys van gebruik.
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.
No: premium external models such as GPT-5, Claude and Gemini are proprietary, carry their own licensing terms, run on purchased tokens and are not part of any comp grant. Only the self-hosted models listed on this page are open source.