Free access by model
Every AI model you can use free, filterable by vision, tool calling, open weights, context and family, with free sources and rate limits per model.
184 models · 68 with two or more free sources or a self-host path
11 of 184 models match. Clear filters
| Model | Family | Context | Vision | Tools | JSON mode | Open weights | Free sources |
|---|---|---|---|---|---|---|---|
| Bulbul v3 | Text to speech | — | — | — | — | — | 2 |
| Eleven v3 | Text to speech | — | — | — | — | — | 2 |
| Flash v2.5 | Text to speech | — | — | — | — | — | 2 |
| Multilingual v2 | Text to speech | — | — | — | — | — | 2 |
| Aura-2 | Text to speech | — | — | — | — | — | 1 |
| Fish Audio S2.1 Pro Free | Text to speech | — | — | — | — | — | 1 |
| Generative voices | Text to speech | — | — | — | — | — | 1 |
| Neural voices | Text to speech | — | — | — | — | — | 1 |
| Octave | Text to speech | — | — | — | — | — | 1 |
| Sonic-3.6 | Text to speech | — | — | — | — | — | 1 |
| Standard voices | Text to speech | — | — | — | — | — | 1 |
Models come from each offer's documented model list and are re-synced nightly; a model page lists every listed offer that serves it free with the limits the provider publishes. A dash means the sources do not say or disagree.