I’m trying to use the groq moonshotai/kimi-k2-instruct model. According to the documentation on moonshotai/Kimi-K2-Instruct · Hugging Face this model isnt’ compatible with OpenAI or Gemini tokenizers and it appears to use it’s own custom tokenizer. Is it possible to configure discourse to use a cus…

TL;DR pick the closest tokenizer and set the maximum context to some thousands less to allow for the difference to not affect you.

Thanks. So I decided to enroll the services of ChatGPT, Gemini and Grok to help me decided which tokenizer to use; which would be the closest match to the Kimi Instruct TikToken/BPE tokenizer to generate the most accurate output from the model. I must say modern AI models are fairly representative …

Also worth noting … majority of coding agents these days don’t even bother with an accurate tokenizer like Discourse does. They just estimate at 4 letters per token. cl100k will be plenty fine for the vast majority of use cases on llms with slightly different tokenizers.

Cómo usar modelos de IA con tokenizadores personalizados

Soporte

Falco (Falco) 14 Diciembre, 2025 00:02 2

TL;DR elige el tokenizador más cercano y establece el contexto máximo en unos miles menos para permitir que la diferencia no te afecte.

Tema		Respuestas	Vistas
Adding Semantic Search feature for our self-hosted discourse site Support ai , ai-search	9	272	19 Marzo 2025
Frustrations on AI spam detector Support ai , spam	8	188	21 Noviembre 2025
AI exceeds LLM token thresholds randomly and unpredictably Support ai	3	110	6 Mayo 2026
Configuring OpenRouter language models Integrations ai	0	1241	10 Diciembre 2024
How to implement Mistral with Embeddings Support ai , related-topics	5	279	11 Abril 2025

Cómo usar modelos de IA con tokenizadores personalizados

Temas relacionados