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% ngā tauira pūtake-whakahaere AI
He pūtake tūwhera ia tauira i whakatūria ki runga i te Free.ai i raro i ngā whakaaetanga whakaaetanga (Apache 2.0, MIT, BSD). Kāore he whakaaetanga Llama/Meta. E whakapono ana tātau me mārama te AI, ka taea te whakamātautau, me te wātea ki te whakamahi.
Chat ngā tauira reo
| Kāhua | Ka whakamahia te take | Whakawhiwhinga | Ka taea te whakataki te pūtake |
|---|---|---|---|
| Qwen2.5-72B | Te kōrero whānui, te whakaaro, te tātaritanga | Apache 2.0 | HuggingFace |
| Qwen2.5-7B | Āhua kōrero tere, mahi māmā | Apache 2.0 | HuggingFace |
| Mistral-7B | Āhua kōrero tere, take ahuwhānui | Apache 2.0 | HuggingFace |
| Phi-3 | He take tika, he pāngarau | MIT | HuggingFace |
Kāhua waehere
| Kāhua | Ka whakamahia te take | Whakawhiwhinga | Ka taea te whakataki te pūtake |
|---|---|---|---|
| Qwen2.5-Coder-32B | Ko te whakatūnga waehere, te whakarerekētanga | Apache 2.0 | HuggingFace |
| DeepSeek-Encoder-V2-Lite | Whakaputanga waehere, arotakenga | MIT | HuggingFace |
Tauira pikitia me te vitio
| Kāhua | Ka whakamahia te take | Whakawhiwhinga | Ka taea te whakataki te pūtake |
|---|---|---|---|
| FLUX.1-te tere | He tere te whakatūnga o te whakaahua | Apache 2.0 | HuggingFace |
| Pāpāho XL tūturu | Ko te whakatūnga whakaahua o te āhua nui | OpenRAIL++ | HuggingFace |
| Kandinsky 2.2 | Ko te whakatūnga o te whakaahua ā-ariā | Apache 2.0 | HuggingFace |
| CogVideoX-2B | Ko te whakatūnga o te ataata mai i te kupu | Apache 2.0 | HuggingFace |
Kāhua kōrero me te oro
| Kāhua | Ka whakamahia te take | Whakawhiwhinga | Ka taea te whakataki te pūtake |
|---|---|---|---|
| Kokoro | Whakatuhi-ki-te-kōrero | Apache 2.0 | HuggingFace |
| Piper | Whakatuhi-ki-te-kōrero | MIT | GitHub |
| MeloTTS | Whakatuhi-ki-te-kōrero | MIT | GitHub |
| Chatterbox | Whakatuhi-ki-te-kōrero | MIT | HuggingFace |
| TTS-Out | Whakatuhi-ki-te-kōrero | Apache 2.0 | HuggingFace |
| faster-whisper | Whakawhitinga ki te kupu | MIT | GitHub |
| Te rākau | Whakawhitinga ki te kupu | Apache 2.0 | GitHub |
| ACE-Step 3.5B | Ko te whakatūnga pūoro | Apache 2.0 | HuggingFace |
| Mel-Band RoFormer | Te wehe pūoro (whakaahua, pūoro) | MIT | arXiv |
| SCNet | Te wehe pūoro (rohe, pāpāho, ētahi atu) | Apache 2.0 | arXiv |
Ka whakamāoritia, te tirohanga me ētahi atu tauira
| Kāhua | Ka whakamahia te take | Whakawhiwhinga | Ka taea te whakataki te pūtake |
|---|---|---|---|
| MadLAD-400 3B | Whakamāoritanga (450+ reo) | Apache 2.0 | HuggingFace |
| Real-ESRGAN | Whakatauiratanga whakaahua | BSD | GitHub |
| BRIA RMBG 2.0 | Whakakore i te papamuri | Apache 2.0 | HuggingFace |
| US | Whakawatanga whakaahua | Apache 2.0 | GitHub |
| Kāore i te whakatata | Whakawhanaketanga o te whakaahua | MIT | GitHub |
| MiniLM-L6-v2 | Ko ngā whakatakituhi kupu | Apache 2.0 | HuggingFace |
| Tetere | OCR | Apache 2.0 | GitHub |
Ka whakawātea te pūtake o te papatono whakahaere
Ka taea e ngā kaiwhakaū o ngā kaupapa pūtake tūwhera e whakahaeretia ana te tono mō tētahi whakawhiwhinga wātea o ngā tohu 15,000,000, te whakawhiwhinga tohu o tētahi tau o Pro, tae atu ki ngā kī API i te taumata tono Pro (100,000 tono i ia marama). Ka taupoki te whakawhiwhinga i ia utauta e haere ana i runga i a tātau tauira whakanoho-i-te-whāiti. Kāore ngā tauira waho Premium i te wātea o te whakawhiwhinga: ka whakamahia e rātau ngā tohu hoko.
Ko wai e whai āhei ana
- E pupuri ana koe i te kaupapa (kaipupuri, kaiwhakakōrero rānei), ā, he tūmatanui tōna pūtahi i raro i tētahi whakaaetanga pūtake tūwhera i whakaaetia e te OSI
- Kua mau tonu, me ngā whakawhiwhinga, whakawhiwhinga rānei i roto i ngā rā o mua o te 90
- Ko ētahi atu tāngata e whakamahi ana i a ia, e whakaaturia ana e tētahi o te iti rawa o: 50 ngā whetū GitHub, 1,000 ngā tāpiritanga pakiaka i tētahi marama, i ngā kaupapa rānei e whakawhirinaki ana ki a ia.
- Ko tētahi whakawhiwhinga ki ia pūkete Free.ai puta noa i te pūtake tūwhera, te kore hua me ngā papatono tīmatanga, i tāpirihia ki tētahi pūkete kua hangaia e koe
Hei tono, tuku i te pātahitanga puna me tētahi pātahitanga ki te whakaaturanga o te whakamahinga (stars, download statistics, dependents or) mai i te āhua o te imeli i runga i tōmu pūkete Free.ai. Ka hoatu mā te tātai hoapā
FAQ
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.
He whakawhiwhinga wātea o ngā tohu 15,000,000 tae atu ki ngā kī API i te taumata tono Pro (100,000 tono i ia marama) mō ngā kaihoahoa o ngā kaupapa pūtake tūwhera i whakawhiwhia ki te whakaaturanga o te whakamahi. E āpiti ana i ngā utauta e haere ana i runga i a tātau tauira whakanoho-i-te-whakahonohono. Kāore ngā tauira Premium o waho i te wāhanga o te whakawhiwhinga, ā, ka whakamahia ngā tohu hoko. Ka tono me tētahi pātahitanga ki tōna puna me te whakaaturanga o te whakamahi.
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.
Kāore: he whakahauhau ngā tauira o waho pēnei i te GPT-5, Claude me te Gemini he tūtohu, e mau ana i ō rātou ake whakaritenga whakaaetanga, e haere ana i ngā tohu hoko, ā, kāore he wāhanga o tētahi whakawhiwhinga. Ko ngā tauira whakawhiwhinga anake i whakarārangitia i tēnei pātaka he pūtake tūwhera.