Prompt Chain

Compare Open Source LLMs on Key Tasks

Promptfoo example comparing DeepSeek, Mistral, Llama, and Qwen on factual tasks through a single OpenRouter key.

Works with githubopenrouter

Maintainer of this project? Claim this page to edit the listing.


81
Spark score
out of 100
Updated 28 days ago
Version code-scan-action-0.1

Add to Favorites

Why it matters

Evaluate and compare the performance of leading open-source large language models (LLMs) like DeepSeek, Mistral, Llama, and Qwen. This asset helps you understand their capabilities across various factual tasks.

Outcomes

What it gets done

01

Benchmark LLM performance on factual assertions.

02

Compare outputs from multiple open-source models.

03

Utilize OpenRouter for model access and comparison.

Install

Add it to your toolbox

Run in your project directory:

curl -fsSL https://spark.entire.vc/get/pfoo-compare-open-source-models | bash

Steps

Steps in the chain

01
Initialize example project
02
Navigate to project directory
03
Set OpenRouter API key
04
Run evaluation
05
View results

Overview

Compare Open Source Models

This example compares DeepSeek, Mistral, Llama, and Qwen on factual accuracy through a single OpenRouter API key. Use it when comparing DeepSeek, Mistral, Llama, and Qwen on factual accuracy without configuring separate provider API keys.

What it does

This example compares DeepSeek, Mistral, Llama, and Qwen on various tasks using factual assertions, accessed through a single OpenRouter API key.

When to use - and when NOT to

Use it when you want a working comparison of four open-source model families on factual accuracy, without configuring separate API keys for each provider. Do not use it if you need models outside this set - the example is specifically scoped to DeepSeek, Mistral, Llama, and Qwen.

Inputs and outputs

Requires OPENROUTER_API_KEY. Running the evaluation scores each model's responses against factual assertions defined in the config, viewable via promptfoo view.

Integrations

Set up via npx promptfoo@latest init --example compare-open-source-models, then npx promptfoo@latest eval and npx promptfoo@latest view. A companion guide covers the comparison methodology in more depth.

Who it's for

Teams evaluating open-source model families - DeepSeek, Mistral, Llama, and Qwen - on factual accuracy through a single OpenRouter key rather than juggling separate provider accounts.

Source README

compare-open-source-models (Open-Source Model Comparison)

This example compares DeepSeek, Mistral, Llama, and Qwen on various tasks with factual assertions using OpenRouter.

See the guide for more details.

Setup

npx promptfoo@latest init --example compare-open-source-models
cd compare-open-source-models

Set the required environment variable:

export OPENROUTER_API_KEY=your-openrouter-key

Run

npx promptfoo@latest eval

View

npx promptfoo@latest view

FAQ

Common questions

Discussion

Questions & comments · 0

Sign In Sign in to leave a comment.