Free/Cheap LLM APIs
(for SillyTavern)
Platform | Platform Type | Deal Type | Cost | Models | Rate limit | Cons | Guide |
---|---|---|---|---|---|---|---|
Google Vertex | Provider | Sign up | Free 300$ for 3 months | Gemini models | — | — | Supported |
Amazon Web Services | Provider | Sign up | Free 200$ for 6 months | Claude models | — | Long setup | Rentry |
Parasail | Provider | Sign up | Free 10$ | Deepseek models, Kimi K2, GLM 4.5 | — | — | Go down |
Sambanova | Provider | Sign up | Free 5$ | Deepseek models | 60 RPD | — | Go down |
Groq | Provider | Sign up | Free Tier | Kimi K2 | 10K TPM | Super low rate limits | Supported |
NVIDIA NIM | Provider | Sign up | Free Tier | Deepseek models, Kimi K2, Qwen 3 | 40RPM (?) | Queue system | |
Routeway | Provider | Waitlist | Free tier | R1T2 Chimera, Kimi K2, GLM 4.5, GPT-4.1 | 20 RPM & 200 RPD | Early development | Go down |
Chutes | Provider | Subscription | 3$/month | Deepseek models, Kimi K2, Qwen 3, GLM 4.5 | 300 RPD | — | Supported |
4risu (Discord) | Proxy | Subscription | Free/Pro tier | Sonnet 3.5/3.7, GPT-4o, GPT-4.1, Deepseek V3.1 | — | Rate limit errors on free tier (?) | Read the #rules on DC |
Anas Proxy (Discord) | Proxy | Subscription | $1.60/day | Most RP models | Unlimited | Parameters/Images not supported | Rentry |
NanoGPT | Gateway | Subscription | 8$/month | Deepseek models, Kimi K2, Qwen 3 | 2000 RPD | — | Supported |
Openrouter | Gateway | One time deposit | Deposit 10$ once | Deepseek models, Kimi K2, Qwen 3 | 1000 RPD | Rate limit errors | Supported |
LLMGateway | Gateway | One time deposit | Deposit 5$ once | R1T2 Chimera, Kimi K2, GLM 4.5, GPT-4.1 | 20 RPM | Some providers are in early development | Go down |
Vercel | Gateway | Sign up | Free 5$ credits (monthly) | Most RP models (except Opus) | — | Credit card needed | Go down |
Atlas Cloud | Gateway | Sign up | Free 0.1$ credits (daily) | Most RP models | — | — | Go down |
Note: ? means "not sure"
How To Use Custom APIs?
- Go to Top Bar → Connection Profile
- Choose Chat Completion → Custom (OpenAI-compatible)
- Enter Base URL and API key
- Click connect and choose a model
Base URLs:
Platform | Base URL |
---|---|
Parasail | https://api.saas.parasail.io/v1 |
Sambanova | https://api.sambanova.ai/v1 |
NVIDIA | https://integrate.api.nvidia.com/v1 |
Routeway | https://api.routeway.ai/v1 |
4risu | https://4risu.org/v1 |
LLMGateway | https://api.llmgateway.io/v1 |
Vercel | https://ai-gateway.vercel.sh/v1 |
Atlas Cloud | https://api.atlascloud.ai/v1 |