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% Amamodeli we-AI avulekile
Yonke imodeli ehlala yona ku-Free.ai ivela emthonjeni ovulekile ngaphansi kwelayisense evumelayo (Apache 2.0, MIT, BSD). Akukho Llama/Meta ilayisense. Sikholelwa ukuthi i-AI kufanele ibe mhlophe, ihlolwe, futhi ikhululekile ukusebenzisa.
Chat Izimo zesilimi
| Imodeli | Sebenzisa isinqumo | Ilayisense | Umsuka |
|---|---|---|---|
| Qwen2.5-72B | Ukukhuluma nsuku zonke, ukucabanga, ukulinganisa | Apache 2.0 | HuggingFace |
| Qwen2.5-7B | Ukuxhumana okukhawulelwe, imisebenzi elula | Apache 2.0 | HuggingFace |
| Mistral-7B | Ukuxhumana okukhawulelayo, okujwayelekile | Apache 2.0 | HuggingFace |
| Phi-3 | Ukuqonda okunembile, i-matematics | MIT | HuggingFace |
Imodeli yekhodi
| Imodeli | Sebenzisa isinqumo | Ilayisense | Umsuka |
|---|---|---|---|
| Qwen2.5-Coder-32B | Ukukhishwa kwekhodi, ukuxazululwa kwephutha | Apache 2.0 | HuggingFace |
| DeepSeek-Coder-V2-Lite | Ukuqedela ikhodi, ukuhlola | MIT | HuggingFace |
Isithombe nevidiyo
| Imodeli | Sebenzisa isinqumo | Ilayisense | Umsuka |
|---|---|---|---|
| FLUX.1-fast | Ukukhiqizwa kwesithombe esisheshayo | Apache 2.0 | HuggingFace |
| Stable Diffusion XL | Ukukhiqizwa kwesithombe esisezingeni eliphakeme | OpenRAIL++ | HuggingFace |
| Kandinsky 2.2 | Ukukhiqizwa kwesithombe esizimele | Apache 2.0 | HuggingFace |
| CogVideoX-2B | Ukukhiqizwa kwevidiyo kusuka kumbhalo | Apache 2.0 | HuggingFace |
Ukukhuluma noMsindo
| Imodeli | Sebenzisa isinqumo | Ilayisense | Umsuka |
|---|---|---|---|
| Kokoro | Umbhalo-kuya-kwezwi | Apache 2.0 | HuggingFace |
| Piper | Umbhalo-kuya-kwezwi | MIT | GitHub |
| MeloTTS | Umbhalo-kuya-kwezwi | MIT | GitHub |
| Chatterbox | Umbhalo-kuya-kwezwi | MIT | HuggingFace |
| OutTTS | Umbhalo-kuya-kwezwi | Apache 2.0 | HuggingFace |
| faster-whisper | Ukukhuluma-ngokubhaliwe | MIT | GitHub |
| I-wax | Ukukhuluma-ngokubhaliwe | Apache 2.0 | GitHub |
| ACE-Step 3.5B | Ukukhishwa komculo | Apache 2.0 | HuggingFace |
| Mel-Band RoFormer | Ukuhlukaniswa komculo (ama-vocals, ama-instrumental) | MIT | arXiv |
| SCNet | Ukuhlukaniswa komculo (ama-drums, ama-bass, abanye) | Apache 2.0 | arXiv |
Ukuhumusha, Umbono & Okunye
| Imodeli | Sebenzisa isinqumo | Ilayisense | Umsuka |
|---|---|---|---|
| I-MadLAD-400 3B | Ukuhumusha (izinhlamvu ezingaphezu kuka-450) | Apache 2.0 | HuggingFace |
| Real-ESRGAN | Ukuphakama kwesithombe | BSD | GitHub |
| BRIA RMBG 2.0 | Ukususa isizinda | Apache 2.0 | HuggingFace |
| i-United States | Ukuhlukaniswa kwesithombe | Apache 2.0 | GitHub |
| Ukulungisa | Umbala wesithombe | MIT | GitHub |
| all-MiniLM-L6-v2 | Ukungenisa umbhalo | Apache 2.0 | HuggingFace |
| I-Tesseract | OCR | Apache 2.0 | GitHub |
I-Open Source Maintainer Program
Abalondolozi bezinhlelo ezivulekile ezigcinwe kahle bangafaka isicelo sokuba nesibopho esimahhala se-15,000,000 tokens, i-token allowance yeminyaka emibili ye-Pro, kanye ne-API keys ku-Pro request tier (100,000 requests ngenyanga). Isibopho sifaka yonke into esebenza kumamodeli ethu ahlala e-self. Amamodeli angaphandle aphezulu angeke abe ingxenye yesibopho: basebenzisa ama-tokens athengiswe.
Obani abafanele
- Ugcina iphrojekthi (umnikazi noma umphathi ojwayelekile), futhi i-repository yayo ivela kwabanye ngaphansi kwelayisense yomthombo ovulekile egunyazwe yi-OSI
- Kugcinwa ngokuqhubekayo, ngezinkontileka noma izikhululo ezidlulele ezinsukwini ezingu-90
- Abanye abantu basebenzisa le phrojekthi, kuboniswe yisinye sezinto ezilandelayo: ama-stars angama-50 eGitHub, amaphakheji angama-1,000 alayishwe inyanga ngayinye, noma amaphrojekthi axhomekeke kuyo
- Usizo olulodwa nge-akhawunti ye-Free.ai esuka kumthombo ovulekile, ongekho emthethweni kanye nezinhlelo zokuqalisa, kufakwe ku-akhawunti oyenze ngaphambili
Ukucela, thumela i-link yendawo yokugcina kanye ne-link yesibonelo sokusetshenziswa (ama-stars, ukukhishwa kwedatha noma okuhlobene) kusuka kukheli le-imeyli le-Free.ai akhawunti yakho. Sebenzisa ngefomu lokuxhumana
Imibuzo ebuzwa kaningi
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.
Usizo olumahhala lwe-15,000,000 tokens plus API keys ku-Pro request tier (100,000 requests per month) for maintainers of actively maintained open source projects with evidence of use. It covers the tools that run on our self-hosted models. Premium external models are not part of the grant and use purchased tokens. Apply with a link to your repository and to the evidence of use.
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.
Akuna: amamodeli angaphandle aphezulu njenge GPT-5, Claude ne Gemini asemthethweni, aphethe izimo zabo zokugunyazwa, aqhube nge-token ethengiswe futhi ayikho ingxenye yosizo lwe-comp. Amamodeli abekwe ngokwezifiso kuphela abhalwe kule khasi avulekile.