Model Comparison

GPT-5 CodexMistral Medium 3
Let's see who can write code better - GPT-5 Codex or Mistral Medium 3

Model Information

GPT-5 Codex

OpenAI

Description

Flagship model from OpenAI optimised for coding tasks

Specifications

Context Window272K
Max Output33K
ReleasedSep 2025
API

Pricing (per 1M)

In / Out
$1.25/ $10

Key Features

CodingAdvanced reasoningMulti-modal native

Strengths

  • State-of-the-art performance
  • Optimised for coding
  • Large context window

Weaknesses

  • Likely very expensive
  • Unknown specifications

Agent Support

No known agents

Mistral Medium 3

Mistral AI

Description

Balanced performance and cost model

Specifications

Context Window128K
Max Output4K
ReleasedApr 2024
API

Pricing (per 1M)

In / Out
$0.4/ $2

Key Features

Balanced performanceGood reasoningCode generationMulti-language support

Strengths

  • Good price-performance ratio
  • Better than small for complex tasks
  • Reliable API
  • Decent code generation

Weaknesses

  • More expensive than small
  • Less capable than large
  • Limited advanced features

Agent Support

No known agents