01
Start with the task
Filter for coding, writing, images, research, chat, or automation before comparing names.
Open-source decision guide
Pick the model that fits the work.Not the loudest launch.
AI Model Picker is a community-maintained catalog for comparing AI models by task, cost, API access, local usage, language support, and privacy.
Current early MVP
- sample models
- 8
- task categories
- 6
- local options
- 5
- fake metrics
- 0
The sample dataset intentionally favors honest caveats over false certainty. Entries marked needs review are invitations to contribute.
A calmer way to choose
Useful filters. Visible uncertainty.
02
See practical trade-offs
Compare price, API access, local usage, Russian support, and privacy in one place.
03
Treat claims as data
Every entry has sources, a verification date, and an explicit review status.
A few starting points
Sample catalog entries
Meta
Llama 3.1 8B Instruct
A compact instruction-tuned language model that can run locally on suitable consumer hardware.
ChatWritingCodingAutomation
Alibaba Cloud
Qwen2.5-Coder 32B Instruct
A code-focused language model with downloadable weights for local development workflows.
CodingAutomation
Mistral AI
Mistral Small 3.1
A downloadable multimodal model intended for conversational assistance and local deployment.
ChatWritingCodingAutomation
Roadmap
What comes after the first useful version.
Trustworthy data comes first. Richer decision tools follow once the contributor workflow is working well.
Discuss the roadmap- 1Verify and expand the starter catalog
- 2Add model detail pages and richer source metadata
- 3Create shareable comparisons and filter presets
- 4Add localization, starting with Russian
Built in public
Know a model better than the dataset does?
Add a source, correct a claim, or propose a new entry. The contribution path is deliberately simple.