Updated weekly · checked
Free LLM APIs right now
19 models you can call for $0 per token today, newest first. Each links to its model page.
| Model | Context | Added | Tool calls |
|---|---|---|---|
| Recursive: Flintrecursive/flint-v1.0:free | 262K ctx | 2026-10-21 | yes |
| XingChen: Xing4.0 29B A4Bxingchen-agi/xing4.0-29b-a4b:free | 262K ctx | 2026-10-16 | yes |
| NVIDIA: Nemotron 3.5 Super VLnvidia/nemotron-3.5-super-vl:free | 1.0M ctx | 2026-10-15 | yes |
| Dots Studio: Dots3-Note Previewdots-studio/dots-3-note-preview:free | 512K ctx | 2026-08-14 | yes |
| LiquidAI: LFM2.5-2.6Bliquid/lfm-2.5-2.6b:free | 65K ctx | 2026-08-11 | yes |
| NVIDIA: Nemotron 3.5 Lightningnvidia/nemotron-3.5-lightning:free | 1.0M ctx | 2026-08-11 | yes |
| Thinking Machines: Inkling Smallthinkingmachines/inkling-small:free | 1.0M ctx | 2026-07-30 | yes |
| Poolside: Laguna S 2.1poolside/laguna-s-2.1:free | 262K ctx | 2026-07-21 | yes |
| Thinking Machines: Inklingthinkingmachines/inkling:free | 1.0M ctx | 2026-07-17 | yes |
| Poolside: Laguna XS 2.1poolside/laguna-xs-2.1:free | 262K ctx | 2026-07-02 | yes |
| Cohere: North Mini Codecohere/north-mini-code:free | 256K ctx | 2026-06-17 | yes |
| NVIDIA: Nemotron 3.5 Content Safetynvidia/nemotron-3.5-content-safety:free | 128K ctx | 2026-06-04 | no |
| NVIDIA: Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55b:free | 1.0M ctx | 2026-06-04 | yes |
| NVIDIA: Nemotron 3 Nano Omninvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free | 256K ctx | 2026-04-28 | yes |
| Google: Gemma 4 26B A4B google/gemma-4-26b-a4b-it:free | 262K ctx | 2026-04-03 | yes |
| Google: Gemma 4 31Bgoogle/gemma-4-31b-it:free | 262K ctx | 2026-04-02 | yes |
| Google: Lyria 3 Pro Previewgoogle/lyria-3-pro-preview | 1.0M ctx | 2026-03-30 | no |
| Google: Lyria 3 Clip Previewgoogle/lyria-3-clip-preview | 1.0M ctx | 2026-03-30 | no |
| Free Models Routeropenrouter/free | 200K ctx | 2026-02-01 | yes |
Price on OpenRouter: $0 per input and output token (the :free variant). Free variants are rate-limited and can end without notice. Other free routes (free tiers inside coding tools, vendor promos) are added only after we check the official page.
Source: OpenRouter models API · checked
FAQ
Which LLM APIs are free right now?
As of 2026-10-28, OpenRouter lists 19 models at $0 per input and output token. The newest is Recursive: Flint, added 2026-10-21.
Which free model has the longest context window?
Thinking Machines: Inkling Small: 1,048,576 tokens.
Which free models support tool calling?
16 of 19: Recursive: Flint, XingChen: Xing4.0 29B A4B, NVIDIA: Nemotron 3.5 Super VL, Dots Studio: Dots3-Note Preview, LiquidAI: LFM2.5-2.6B, NVIDIA: Nemotron 3.5 Lightning, Thinking Machines: Inkling Small, Poolside: Laguna S 2.1, Thinking Machines: Inkling, Poolside: Laguna XS 2.1, Cohere: North Mini Code, NVIDIA: Nemotron 3 Ultra, NVIDIA: Nemotron 3 Nano Omni, Google: Gemma 4 26B A4B , Google: Gemma 4 31B, Free Models Router.
How do I call a free model?
Use the model id with the ':free' suffix (shown under each name) on OpenRouter's OpenAI-compatible API at https://openrouter.ai/api/v1.
What's the catch?
Free variants are rate-limited, and a model can stop being free without notice. Check the model's page before you build on it. This list is refreshed weekly.
Same data as the Agent Stack Index (JSON). More: Cheapest DeepSeek V4.1 · OpenRouter alternatives