AI · List
Top Open-Source Models by Use Case (Jan 2026)
A cheat sheet mapping open-source AI models to specific tasks.
This reference list maps standout open-source models to specific tasks including coding, OCR, reasoning, and image generation. It helps developers and self-hosters pick the right open model for a job without sifting through benchmarks. It reflects model recommendations as of early 2026.
The list
> OCR — qwen 3-vl, glm 4.6v
> cheap coding — qwen 3 coder
> coding — glm 4.7, minimax m2.1
> reasoning — deepseek speciale
> writing — kimi k2, kimi k2 thinking
> general purpose — deepseek v3.2
> image gen — flux 2 dev, z-image-turbo
> image editing — qwen-image-edit-2509
What it's for
- Select an open-source model tailored to specific tasks like OCR, coding, or image editing
- Find lightweight or budget-friendly coding models for local deployment
- Identify current open-source alternatives to proprietary LLMs for reasoning and writing
Key points
- Categorized picks across OCR, coding, reasoning, writing, general purpose, and image workflows
- Distinguishes between cheap coding options and full-power coding models
- Highlights specific modern models including Qwen, DeepSeek, GLM, Kimi, and Flux variants
- Provides a fast starting point for evaluation without reading benchmark papers
Pricing
Free reference list; models listed are open-source weights
Details
- Type
- List
- Category
- AI
- Topic
- Model Selection
- Open source
- Yes
- Platforms
- Self-hosted, API
- For
- Developers, AI engineers, and self-hosters choosing open-weights models
- Alternative to
- LMSYS Chatbot Arena, Hugging Face Open LLM Leaderboard
- Tags
- open source, llm, cheat sheet, reference, ai models, self hosted
Why it was saved
- @askOkara: Quick reference for choosing an open-source model per use case without benchmark research.
Related
Browser that installs and runs open-source AI apps with one click
A comprehensive guide to prompt engineering techniques and LLM parameters by Google.
Open-weights fine-tune of Llama 3.1 built for agentic function calling.
Private, on-device AI assistant to chat with documents and local models.
Open-source platform and framework for building autonomous AI agents
Open-weights AI models and chat interface for reasoning, coding, and research.
Updated