Llm mistralai

Overview

Mistral.ai is a company that builds OS language models, but also offers a platform hosting those models. You can use Instructor with Mistral API by configuring the client as demonstrated below. Please note that the larger Mistral models support native JSON object response_format, which is much more reliable than Instructor markdown-JSON fallback. Inference feature compatibility:

tool calling - supported (Mistral-Small / Mistral-Medium / Mistral-Large)
native JSON object response_format - recommended (Mistral-Small / Mistral-Medium / Mistral-Large)
Instructor markdown-JSON fallback - fallback mode (Mistral 7B / Mixtral 8x7B)

Example

<?php

use Cognesy\Messages\Messages;
use Cognesy\Polyglot\Inference\Inference;
use Cognesy\Utils\Str;

require 'examples/boot.php';

$answer = Inference::using('mistral')
    ->with(
        messages: Messages::fromString('What is the capital of France'),
        options: ['max_tokens' => 256]
    )
    ->get();

echo "USER: What is capital of France\n";
echo "ASSISTANT: $answer\n";
assert(Str::contains($answer, 'Paris'));
?>

A01_Basics

A02_Advanced

A03_Troubleshooting

A04_APISupport

A05_Extras

B01_LLM

B02_LLMAdvanced

B03_LLMTroubleshooting

B04_LLMApiSupport

B05_LLMExtras

C01_Http

D01_Agents

D02_AgentBuilder

D03_AgentTemplates

D04_AgentSessions

D05_AgentTroubleshooting

D10_AgentCtrl

D20_Sandbox

D21_SandboxAPIs

Z01_ZeroShot

Z02_FewShot

Z03_ThoughtGen

Z04_Ensembling

Z05_SelfCriticism

Z06_Decomposition

Z07_Misc

Overview

Example

​Overview

​Example

Overview

Example