MythoMax 13B logo

MythoMax 13B

Visit

MythoMax-L2-13b is Gryphe's Llama 2 13B merge of MythoLogic-L2 and Huginn for roleplay and storywriting, with an Alpaca-style prompt and a legacy label.

Share:
View alternatives

MythoMax 13B (Gryphe/MythoMax-L2-13b) is a Llama 2-based merge created by Gryphe and published on Hugging Face. The model card describes it as an improved version of the earlier MythoMix, built from MythoLogic-L2-13b and Huginn-13b-FP16 with an experimental tensor-type merge technique, and as proficient at both roleplaying and storywriting. The card places it in the Mytho Series (Llama 2) collection, which it marks as legacy, and notes that Llama 3 has since been released.

Key features

  • Merge recipe: per the card, MythoLogic-L2 acts as the input side and Huginn as the output side, aimed at writing ability.
  • Prompt format: Alpaca style, with a system prompt or character card, then ### Instruction: and ### Response: sections.
  • Quantized builds: TheBloke provides GGUF, GPTQ and AWQ versions, and the Hub lists 18 quantized models that work with llama.cpp, Ollama, LM Studio and similar apps.
  • License: listed as "other", so check the terms on the model page.
  • Status: part of a collection that the card marks as legacy; the card gives no context length.

Use cases

  • Roleplay and character chat on local hardware.
  • Storywriting and creative drafts.
  • Running a small 13B model through llama.cpp, Ollama or LM Studio.

Pricing and availability

The weights are free to download from Hugging Face; the license is listed as "other". Running a 13B model locally is practical with the quantized builds. When we checked on 2026-10-29 the card showed 410 likes and about 1,459 downloads in the last month. No official successor is named on the card, but its legacy label and the Llama 3 note suggest looking at newer models.

Getting started

  1. Pick a quantized build such as a GGUF file that fits your memory.
  2. Load it in llama.cpp, Ollama or LM Studio.
  3. Format prompts in the Alpaca style, with a system prompt or character card.
  4. Compare against a newer model such as Llama 3.1 8B Instruct if quality matters more than style.

FAQ

What is MythoMax 13B?

A Llama 2 13B merge by Gryphe of MythoLogic-L2-13b and Huginn-13b-FP16, aimed at roleplay and storywriting.

What prompt format does it use?

Alpaca style, with ### Instruction: and ### Response: sections.

Is it a current model?

The model card puts it in a Llama 2 collection that it marks as legacy and notes that Llama 3 has been released.

Can I run it locally?

Yes. TheBloke provides GGUF, GPTQ and AWQ builds that work with llama.cpp, Ollama and LM Studio.

Alternatives

Conclusion

MythoMax 13B is a documented legacy Llama 2 merge that still runs well locally for roleplay and storywriting. Read the model card, test it against newer models, and browse more in the models category.

Comments

No comments yet. Be the first to comment!