Which API to use with SillyTavern
Use Outrun in SillyTavern as a Chat Completion source set to Custom (OpenAI-compatible), with the endpoint https://api.outrun.to/v1 and one prepaid key. At 16k tokens of context per turn, 100 turns cost from $0.04 on Claude Haiku 5.5 (2), or from $0.16 on a less-restricted model (GLM-5.3 Flash Uncensored).
Prices as of , from the live price feed.
$0.10free API credit for eligible new accounts.
Cost at 16k context per turn, as in the table below.
How do I connect SillyTavern?
These are the settings from the SillyTavern setup guide, which walks through each step.
API type: Chat Completion
Chat Completion Source: Custom (OpenAI-compatible)
Custom endpoint: https://api.outrun.to/v1
API key: your Outrun key
Model ID: glm-5.3-flash-uncensoredWhich models fit my budget?
General chat models, sorted by the cost of 100 turns at 16k context.
| Model | Input | Output | Cached input | 100 turns at 16k |
|---|---|---|---|---|
Claude Haiku 5.5 (2)claude-haiku-5-5-2 | $0.02 | $0.10 | $0.01 | $0.04 |
Grok 4.7grok-4.7up to 200k tokens | $0.09 | $0.26 | $0.03 | $0.15 |
Claude Haiku 5.5claude-haiku-5-5up to 100k tokens | $0.09 | $0.42 | $0.01 | $0.15 |
GPT-6.1 Solgpt-6.1-solup to 272k tokens | $0.12 | $0.59 | $0.01 | $0.21 |
Claude Sonnet 5.5claude-sonnet-5-5 | $0.49 | $2.45 | $0.03 | $0.86 |
DeepSeek V4 Prodeepseek-v4-pro | $1.35 | $4.05 | $0.05 | $2.29 |
Prices as of , from the live price feed. 100 turns = each turn sends 16,000 input tokens and returns 300 output tokens, with no cache discount. GLM-5.3 Uncensored and GLM-5.3 Flash Uncensored add about 14,000 input tokens of built-in prompt per turn, included here.
Models from the SillyTavern setup guide: claude-sonnet-5-5, deepseek-v4-pro, gpt-6.1-sol. The others are the lowest cost per 100 turns.
Which less-restricted models work with SillyTavern?
Uncensored and abliterated models, sorted the same way. Acceptable-use rules are below.
| Model | Input | Output | Cached input | 100 turns at 16k |
|---|---|---|---|---|
GLM-5.3 Flash Uncensoredglm-5.3-flash-uncensored | $0.05 | $0.05 | $0.01 | $0.16 |
GLM-5.3 Flash Uncensored (2)glm-5.3-flash-uncensored-2 | $0.17 | $0.17 | $0.02 | $0.28 |
GLM-5.3 Uncensoredglm-5.3-uncensored | $0.10 | $0.10 | $0.02 | $0.31 |
DeepSeek V4 Pro Uncensoreddeepseek-v4-pro-uncensored | $0.50 | $0.50 | $0.10 | $0.82 |
MiMo V2.6 Flash Uncensoredmimo-v2.6-flash-uncensored | $1.00 | $3.00 | $0.40 | $1.69 |
glm-5.3-abliteratedglm-5.3-abliterated | $1.40 | $1.40 | $0.40 | $2.29 |
Prices as of , from the live price feed. 100 turns = each turn sends 16,000 input tokens and returns 300 output tokens, with no cache discount. GLM-5.3 Uncensored and GLM-5.3 Flash Uncensored add about 14,000 input tokens of built-in prompt per turn, included here.
Models from the SillyTavern setup guide: deepseek-v4-pro-uncensored, glm-5.3-abliterated, glm-5.3-flash-uncensored, mimo-v2.6-flash-uncensored. The others are the lowest cost per 100 turns.
Which rules apply to roleplay?
The same Terms as every other request.
- You must be 18 or older to use Outrun.
- The Terms apply to every request, including roleplay: no content that exploits children, harasses people or shares private information without authorization.
- Behaviour and quality differ by model, so test the one you choose with a short chat first.
SillyTavern API questions
The setup guide covers context, response length and streaming.
Which API should I use with SillyTavern?
Any OpenAI-compatible API works through the Custom (OpenAI-compatible) Chat Completion source. With Outrun you use one key at https://api.outrun.to/v1 and pick any model ID.
How much do 100 SillyTavern turns cost?
At 16k tokens of context per turn and 300 output tokens, 100 turns cost from $0.04 on Claude Haiku 5.5 (2). Less-restricted models start at $0.16 on GLM-5.3 Flash Uncensored.
Which connection settings do I use?
API type Chat Completion, source Custom (OpenAI-compatible), custom endpoint https://api.outrun.to/v1, then your Outrun key and a model ID. Do not append /chat/completions.
Can I use uncensored models in SillyTavern?
Yes. Uncensored and abliterated models use the same key and endpoint. The Terms still apply, and you must be 18 or older.
Why does context size change the cost?
SillyTavern sends the chat history with every message, so each turn is billed for all the context tokens it sends. A smaller context lowers the cost per turn.
Does Outrun keep my chats?
Outrun does not store your prompts, images or outputs unless you choose to save them.