Pygmalion: Mythalion 13B

Text input Text output
Author's Description

A blend of the new Pygmalion-13b and MythoMax. #merge

Key Specifications
Cost
$$$
Context
4K
Parameters
13B
Released
Sep 01, 2023
Speed
Ability
Reliability
Supported Parameters

This model supports the following parameters:

Stop Presence Penalty Top P Temperature Min P Seed Frequency Penalty Max Tokens
Performance Summary

Pygmalion: Mythalion 13B, a blend of Pygmalion-13b and MythoMax, demonstrates moderate speed performance, ranking in the 32nd percentile across six benchmarks. Its pricing is competitive, positioned in the 47th percentile. The model's overall accuracy across various benchmarks is generally low. It exhibits a notable weakness in complex cognitive tasks such as Reasoning (11th percentile accuracy), Ethics (7th percentile), and General Knowledge (11th percentile), where its performance is significantly below average. Coding (16th percentile) and Instruction Following (20th percentile) also show limited proficiency. A key strength, however, is observed in Email Classification, where it achieves a respectable 91.0% accuracy, placing it in the 20th percentile for this specific task. This suggests a capability for structured classification problems. Despite this singular strength, the model's performance across most other categories indicates a need for substantial improvement in areas requiring deeper understanding, logical inference, and broad factual recall. Its competitive pricing and moderate speed might make it suitable for specific, less cognitively demanding classification applications, but it is not recommended for tasks requiring high accuracy in reasoning, ethics, or general knowledge.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $0.7
Completion $1.1

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
Featherless
Featherless | pygmalionai/mythalion-13b 4K $0.7 / 1M tokens $1.1 / 1M tokens
Parasail
Parasail | pygmalionai/mythalion-13b 4K $0.7 / 1M tokens $1.1 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Free Executions Accuracy Cost Duration