Gemini 3.5 Flash Presets for SillyTavern

Gemini 3.5 Flash Presets for SillyTavern

Gemini 3.5 Flash presets are plentiful. If you’ve used Gemini 3.5 Flash for any length of time, you already know the deal. It’s fast, it’s cheap, and the prose is the best the Gemini line has produced to date. “Less ‘Geminisms'” is the consensus, and that tracks with what I’ve seen in long sessions. The main practical tip is to turn off streaming on AI Studio, otherwise the API-level filter will block your message sends mid-stream. On the censorship question, Flash is less restrictive than previous Gemini models for NSFW, but violence has some filter friction on AI Studio specifically. NanoGPT and OpenRouter tend to be cleaner. The 1M context window and Flash-tier pricing make it a strong value pick.

Three presets that play nice with Flash:

Stab’s EDH – best structured pick

Link to Github Here

Stab’s EDH is explicitly listed as compatible with Gemini 3.0 and carries over to 3.5. Gemini’s strong instruction-following behavior pairs naturally with the EDH’s tiered authority structure, and the Visual Toolkit’s HTML generation features are particularly effective with Gemini’s multimodal output quality. If you want structured character rule enforcement and HTML visual elements in your RP, this is the one to start from. DeepSeek V3.2 Presets

NemoEngine (v10)

Link to NemoEngine here

NemoEngine specifically targets Gemini and DeepSeek as primary models alongside Claude. Its 5.9 update was specifically a Gemini/DeepSeek release. The scratchpad CoT, which tracks character knowledge, emotions, directive adherence, and parallel storylines, is particularly effective with Gemini’s reasoning mode. For Gemini thinking models, navigate to AI Response Formatting, then Reasoning, then activate Auto-Parse, set Prefix to and Suffix to . The NemoPresetExt extension adds dropdown toggle management for all modules. Power-user territory, but worth it if you want HTML visual overlays and modular control.

Marinara’s Universal Preset (v10.0)

Link to Marinara’s here GLM 5.2 Presets

Marinara’s #1 recommended model is Gemini 3.1 Pro, with Flash as a close secondary. The preset’s modular reasoning toggle is directly applicable to Gemini. Marinara recommends disabling reasoning for roleplay across all models, which aligns with the community tip about Flash speed optimization. Lightweight, and works out of the box with no model-specific configuration.

Gemini 3.5 Flash is not the model I reach for when I want the weirdest, most indulgent prose, but it is absolutely the model I keep around for fast daily sessions where cost and context matter. If that sounds like your use case, start with Stab’s and you’re set.

More Presets here.

Leave a Reply

Your email address will not be published. Required fields are marked *