AI Models

Here you can see the LLM models used for the bot responses, ranked by their score. New games automatically use the best-rated available model for the bot.

Your ratings on this page count 3 times in the model score. You can also rate the bot answers directly during your games (they count once).

koboldcpp/L3-Super-Nova-RP-8B

Score: 69.0

koboldcpp/Ministral-3-14B-Reasoning-2512-UD-IQ3_XXS

Score: 20.0

koboldcpp/gemma-4-31B-it-heretic

Score: 16.0

koboldcpp/Gemma-4-E4B-Uncensored-HauhauCS-Aggressive

Score: 11.0

koboldcpp/L3-8B-Stheno-v3.2

Score: 6.0

koboldcpp/L3-8B-Stheno-v3.2-Q8_0

Score: 3.0

koboldcpp/debug-Serenity-31B-v1.1-Q4_K_M

Score: 0.0

koboldcpp/Huihui-Qwen3.5-0.8B-abliterated.Q4_K_M

Score: 0.0

koboldcpp/Qwen/Qwen3.5-0.8B

Score: 0.0

koboldcpp/Magidonia-24B-v4.3

Score: 0.0

koboldcpp/Llama-3.2-3B

Score: 0.0

koboldcpp/Rocinante-X-12B

Score: 0.0

koboldcpp/L3-8B-Stheno-v3.2-IQ3_S-imat

Score: 0.0

koboldcpp/L3-8B-Stheno-v3.2-Q5_K_M

Score: 0.0

koboldcpp/Gemma-4-31B-it-Scotoma-2

Score: 0.0

koboldcpp/L3-8B-Stheno-v3.2-Q4_K_S

Score: 0.0

koboldcpp/Qwen3.5-9B

Score: -1.0

koboldcpp/Behemoth-128B-v3b-Q4_K_M

Score: -1.0

koboldcpp/mini-magnum-12b-v1.1

Score: -1.0

coder3101/gemma-4-E4B-it-qat-q4_0-unquantized-heretic

Score: -1.0

koboldcpp/Angelic_Eclipse-12B

Score: -2.0

aphrodite/TheDrummer/Skyfall-31B-v4.2

Score: -3.0

koboldcpp/Artemis-31B-v1m-Q6_K.gguf

Score: -3.0

koboldcpp/Judas-Uncensored-3.2-1B.Q8

Score: -4.0

koboldcpp/digo-prayudha/unsloth-llama-3.2-1b-gguf

Score: -4.0

koboldcpp/mradermacher/pythia-70m-deduped.f16.gguf

Score: -8.0

koboldcpp/Qwen_Qwen3-0.6B-IQ4_XS

Score: -9.0

koboldcpp/Llama-3.2-1B-Instruct

Score: -10.0

aphrodite/TheDrummer/Cydonia-24B-v4.3

Score: -13.0