Use case

Best free AI models

These models are free to use either via a hosted free tier (look for ':free' suffixes) or because the weights are open and runnable on your own hardware.

Top 25 models for free models

#ModelProviderContextInput price / 1MTier
1OpenAI: gpt-oss-120b (free)OpenAI131K
Long context (128K+)
$0.00Budget
2OpenAI: gpt-oss-20b (free)OpenAI131K
Long context (128K+)
$0.00Budget
3DeepSeek: DeepSeek V4 Flash 0731 (free)DeepSeek1.0M
Ultra context (1M+)
$0.00Budget
4MiniMax: MiniMax M3 (free)Minimax1.0M
Ultra context (1M+)
$0.00Budget
5NVIDIA: Nemotron 3 Ultra (free)Nvidia1M
Ultra context (1M+)
$0.00Budget
6NVIDIA: Nemotron 3.5 Lightning (free)Nvidia1M
Ultra context (1M+)
$0.00Budget
7Qwen: Qwen3.6 Plus Preview (free)Qwen1M
Ultra context (1M+)
$0.00Budget
8Thinking Machines: Inkling (free)Thinkingmachines1.0M
Ultra context (1M+)
$0.00Budget
9Thinking Machines: Inkling Small (free)Thinkingmachines1.0M
Ultra context (1M+)
$0.00Budget
10Apodex: Apodex 1.1 Mini (free)Apodex262K
Long context (128K+)
$0.00Budget
11Cohere: North Mini Code (free)Cohere256K
Long context (128K+)
$0.00Budget
12Dots Studio: Dots3-Note Preview (free)Dots Studio512K
Long context (128K+)
$0.00Budget
13Google: Gemma 4 26B A4B (free)Google262K
Long context (128K+)
$0.00Budget
14Google: Gemma 4 31B (free)Google262K
Long context (128K+)
$0.00Budget
15inclusionAI: Ling 3.0 Flash Fin (free)Inclusionai262K
Long context (128K+)
$0.00Budget
16inclusionAI: Ling 3.0 Flash Sante (free)Inclusionai262K
Long context (128K+)
$0.00Budget
17inclusionAI: Ling 3.0 Flash VL (free)Inclusionai262K
Long context (128K+)
$0.00Budget
18inclusionAI: Ling 3.0 Tiny (free)Inclusionai262K
Long context (128K+)
$0.00Budget
19Ling-3.0-flash (free)Inclusionai262K
Long context (128K+)
$0.00Budget
20Nex AGI: Nex-N2.5-Mini (free)Nex Agi262K
Long context (128K+)
$0.00Budget
21Nex AGI: Nex-N2.5-Pro (free)Nex Agi262K
Long context (128K+)
$0.00Budget
22NVIDIA: Nemotron 3 Nano 30B A3B (free)Nvidia256K
Long context (128K+)
$0.00Budget
23NVIDIA: Nemotron 3 Nano Omni (free)Nvidia256K
Long context (128K+)
$0.00Budget
24NVIDIA: Nemotron 3 Super (free)Nvidia262K
Long context (128K+)
$0.00Budget
25Poolside: Laguna S 2.1 (free)Poolside262K
Long context (128K+)
$0.00Budget

Frequently asked questions

Are these models really free?

Models with ':free' in the id are available at no cost on OpenRouter's free tier (with rate limits). Open-weight models are free if you self-host — you only pay for compute.

Related use cases

Advertisement