cyperx@terminal : ~/models $

> ai models

# tracking the models that power ai coding agents

[6 providers] [22 models]

/timeline

●●● releases.log
[Oct 22, 2025] Anthropic Claude Sonnet 4.5 balanced
[Sep 19, 2025] Qwen Qwen3 frontier
[Jul 23, 2025] Meta Llama 4 Scout fast
[Jul 23, 2025] Meta Llama 4 Maverick frontier
[May 22, 2025] Anthropic Claude Opus 4 frontier
[May 22, 2025] Anthropic Claude Sonnet 4 balanced
[Apr 17, 2025] Google Gemini 2.5 Flash fast
[Apr 16, 2025] OpenAI o3 reasoning
[Apr 16, 2025] OpenAI o4-mini reasoning
[Apr 14, 2025] OpenAI GPT-4.1 frontier
[Mar 25, 2025] Google Gemini 2.5 Pro frontier
[Feb 05, 2025] Google Gemini 2.0 Flash fast
[Dec 17, 2024] OpenAI o1 reasoning
[Dec 06, 2024] Meta Llama 3.3 70B balanced
[Nov 12, 2024] Qwen Qwen 2.5 Coder balanced

/by_provider

●●● anthropic.json

Anthropic

Claude Sonnet 4.5 balanced 200K ctx
released: Oct 22, 2025 id: claude-sonnet-4-5

// Hybrid reasoning model

Claude Opus 4 frontier 200K ctx
released: May 22, 2025 id: claude-opus-4

// Most capable reasoning model

Claude Sonnet 4 balanced 200K ctx
released: May 22, 2025 id: claude-sonnet-4

// Best balance of speed and capability

Claude Haiku 3.5 fast 200K ctx
released: Oct 29, 2024 id: claude-haiku-3-5

// Fastest Claude model

●●● google.json

Google

Gemini 2.5 Flash fast 1M ctx
released: Apr 17, 2025 id: gemini-2.5-flash

// Fast and efficient

Gemini 2.5 Pro frontier 1M ctx
released: Mar 25, 2025 id: gemini-2.5-pro

// Thinking model with massive context

Gemini 2.0 Flash fast 1M ctx
released: Feb 05, 2025 id: gemini-2.0-flash

// Multimodal with tool use

●●● meta.json

Meta

Llama 4 Scout fast 128K ctx
released: Jul 23, 2025 id: llama-4-scout

// Efficient open model

Llama 4 Maverick frontier 128K ctx
released: Jul 23, 2025 id: llama-4-maverick

// Most capable Llama model

Llama 3.3 70B balanced 128K ctx
released: Dec 06, 2024 id: llama-3.3-70b

// Cost-effective performance

●●● mistral.json

Mistral

Mistral Small fast 32K ctx
released: Sep 18, 2024 id: mistral-small

// Cost-effective alternative

Mistral Large 2 frontier 128K ctx
released: Jul 24, 2024 id: mistral-large-2

// Flagship reasoning model

Codestral balanced 32K ctx
released: May 29, 2024 id: codestral

// Code-specialized model

●●● openai.json

OpenAI

o3 reasoning 200K ctx
released: Apr 16, 2025 id: o3

// Latest reasoning model

o4-mini reasoning 200K ctx
released: Apr 16, 2025 id: o4-mini

// Efficient reasoning

GPT-4.1 frontier 1M ctx
released: Apr 14, 2025 id: gpt-4.1

// Coding-optimized

o1 reasoning 200K ctx
released: Dec 17, 2024 id: o1

// Advanced reasoning

GPT-4o mini fast 128K ctx
released: Jul 18, 2024 id: gpt-4o-mini

// Affordable and fast

GPT-4o frontier 128K ctx
released: May 13, 2024 id: gpt-4o

// Multimodal flagship

●●● qwen.json

Qwen

Qwen3 frontier 128K ctx
released: Sep 19, 2025 id: qwen3

// Latest flagship model

Qwen 2.5 Coder balanced 128K ctx
released: Nov 12, 2024 id: qwen-2.5-coder

// Code-optimized variant

Qwen 2.5 balanced 128K ctx
released: Sep 19, 2024 id: qwen-2.5

// Strong coding capabilities