# LLM Model Reference
# See https://llmstxt.org for format details

## Anthropic

- Claude Opus 4.5 (claude-opus-4-5-20251101)
  - Released: 2025-11
  - Context: 200k
  - Input: $5.00/1M
  - Output: $25.00/1M
  - Description: Premium model combining maximum intelligence with practical performance for enterprise workflows.

- Claude Sonnet 4.5 (claude-sonnet-4-5-20250929)
  - Released: 2025-09
  - Context: 200k
  - Input: $3.00/1M
  - Output: $15.00/1M
  - Description: Most intelligent model offering best balance of intelligence, speed, and cost.

- Claude Haiku 4.5 (claude-haiku-4-5-20251001)
  - Released: 2025-10
  - Context: 200k
  - Input: $1.00/1M
  - Output: $5.00/1M
  - Description: Fastest model with near-frontier intelligence at the most economical price point.

## Cohere

- Command A (command-a-03-2025)
  - Released: 2025-03
  - Context: 256k
  - Input: $2.50/1M
  - Output: $10.00/1M
  - Description: Most performant model excelling at tool use, agents, RAG, and multilingual tasks.

- Command A Reasoning (command-a-reasoning-08-2025)
  - Released: 2025-08
  - Context: 256k
  - Input: $2.50/1M
  - Output: $10.00/1M
  - Description: First reasoning model capable of 'thinking' before output generation.

## DeepSeek

- DeepSeek V3.2 Chat (deepseek-chat)
  - Released: 2025-12
  - Context: 128k
  - Input: $0.28/1M
  - Output: $0.42/1M
  - Description: Ultra-low pricing model with 50% cost reduction from previous version.

- DeepSeek V3.2 Reasoner (deepseek-reasoner)
  - Released: 2025-12
  - Context: 128k
  - Input: $0.28/1M
  - Output: $0.42/1M
  - Description: Thinking mode with extended output and reasoning capabilities.

## Google

- Gemini 3 Pro (gemini-3-pro)
  - Released: 2025-11
  - Context: 1,048.576k
  - Input: $2.00/1M
  - Output: $12.00/1M
  - Description: Google's flagship polymath model with 1M context, top benchmark scores in PhD-level reasoning.

- Gemini 3 Deep Think (gemini-3-deep-think)
  - Released: 2025-12
  - Context: 1,048.576k
  - Input: $3.00/1M
  - Output: $18.00/1M
  - Description: Enhanced reasoning model with step-by-step thinking for complex analytical tasks.

- Gemini 2.5 Pro (gemini-2.5-pro)
  - Released: 2025-06
  - Context: 1,048.576k
  - Input: $1.25/1M
  - Output: $10.00/1M
  - Description: State-of-the-art thinking model excelling at complex problems in code, math, and STEM.

- Gemini 2.5 Flash (gemini-2.5-flash)
  - Released: 2025-06
  - Context: 1,048.576k
  - Input: $0.30/1M
  - Output: $2.50/1M
  - Description: Best price-performance model with hybrid reasoning and 1M token context window.

- Gemini 2.0 Flash (gemini-2.0-flash)
  - Released: 2025-02
  - Context: 1,048.576k
  - Input: $0.10/1M
  - Output: $0.40/1M
  - Description: Workhorse model with native tool use and 1M token context.

## Mistral

- Mistral Large 3 (mistral-large-latest)
  - Released: 2025-12
  - Context: 256k
  - Input: $0.50/1M
  - Output: $1.50/1M
  - Description: State-of-the-art open-weight general-purpose multimodal model.

- Codestral (codestral-latest)
  - Released: 2025-07
  - Context: 256k
  - Input: $0.20/1M
  - Output: $0.60/1M
  - Description: Cutting-edge language model optimized specifically for coding tasks.

- Mistral Small 3.2 (mistral-small-latest)
  - Released: 2025-06
  - Context: 128k
  - Input: $0.20/1M
  - Output: $0.60/1M
  - Description: Updated small model balancing performance and efficiency.

## OpenAI

- GPT-5.2 (gpt-5.2)
  - Released: 2025-12
  - Context: 400k
  - Input: $1.75/1M
  - Output: $14.00/1M
  - Description: OpenAI's latest flagship model with enhanced reasoning, coding, and agentic capabilities.

- GPT-5.2 Pro (gpt-5.2-pro)
  - Released: 2025-12
  - Context: 400k
  - Input: $21.00/1M
  - Output: $168.00/1M
  - Description: Most advanced and precise model for high-stakes scenarios and enterprise customers requiring maximum intelligence.

- GPT-5.1 (gpt-5.1)
  - Released: 2025-09
  - Context: 400k
  - Input: $1.25/1M
  - Output: $10.00/1M
  - Description: OpenAI's flagship model for complex reasoning, coding, and agentic tasks with 400K context window.

- GPT-5 Mini (gpt-5-mini)
  - Released: 2025-08
  - Context: 400k
  - Input: $0.25/1M
  - Output: $2.00/1M
  - Description: Cost-effective GPT-5 variant optimized for performance and speed.

- GPT-5.1-Codex (gpt-5.1-codex)
  - Released: 2025-09
  - Context: 400k
  - Input: $1.25/1M
  - Output: $10.00/1M
  - Description: Specialized model tuned for software engineering and agentic coding workflows with 400K context.

- GPT-5.1-Codex-Max (gpt-5.1-codex-max)
  - Released: 2025-09
  - Context: 400k
  - Input: $1.25/1M
  - Output: $10.00/1M
  - Description: Frontier agentic coding model with 400k context, native compaction, and specialized training for long-horizon software engineering.

- GPT-5.1-Codex-Mini (gpt-5.1-codex-mini)
  - Released: 2025-09
  - Context: 400k
  - Input: $0.25/1M
  - Output: $2.00/1M
  - Description: Compact, cost-optimized version targeting lightweight coding workflows like linting and formatting.

- GPT-4.1 (gpt-4.1)
  - Released: 2025-04
  - Context: 1,000k
  - Input: $2.00/1M
  - Output: $8.00/1M
  - Description: Improved instruction following with 1M token context window.

- o3 (o3)
  - Released: 2025-04
  - Context: 200k
  - Input: $2.00/1M
  - Output: $8.00/1M
  - Description: Advanced reasoning model optimized for science, math, and complex analytical tasks.

## xAI

- Grok 4.1 (grok-4.1)
  - Released: 2025-11
  - Context: 256k
  - Input: $3.50/1M
  - Output: $17.50/1M
  - Description: Latest xAI model with high emotional intelligence and strong logical reasoning. Top of LM Arena leaderboard.

- Grok 4 (grok-4-0709)
  - Released: 2025-07
  - Context: 256k
  - Input: $3.00/1M
  - Output: $15.00/1M
  - Description: xAI's flagship reasoning model with 256K context and function calling.

- Grok 4 Fast (grok-4-fast)
  - Released: 2025-07
  - Context: 2,000k
  - Input: $0.20/1M
  - Output: $0.50/1M
  - Description: Fast version with 2M context window for lighter, faster use cases.