Open source AI coding agent. Designed for large projects and real world tasks.
Go aiai-agentsai-developer-toolsai-toolsclicommand-linedeveloper-toolsgitgolanggpt-4llmopenaipolyglot-programmingterminalterminal-basedterminal-ui
Add models overview documentation
Dane Schneidere166562a8d07
diff --git a/docs/docs/models/models-overview.md b/docs/docs/models/models-overview.mdnew file mode 100644index 00000000..3c465caa--- /dev/null+++ b/docs/docs/models/models-overview.md@@ -0,0 +1,111 @@+---+sidebar_position: 1+sidebar_label: Overview+---++# Models Overview++By default, Plandex uses a mix of Anthropic, OpenAI, and Google models. While this is a good starting point, and is the recommended way to use Plandex for most users, full customization of models and providers is also supported.++## Roles and Model Packs++Plandex has multiple [roles](./roles.md) which are responsible for different aspects of planning, coding, and applying changes. Each role can be assigned a different model. A **model pack** is a mapping of roles to specific models.++## Built-in Models and Model Packs++Plandex provides a curated set of built-in models and model packs.++You can see the list of available model packs with:++```bash+\model-packs # REPL+plandex model-packs # CLI+```++You can see the details of a specific model pack with:++```bash+\model-packs show model-pack-name # REPL+plandex model-packs show model-pack-name # CLI+```++You can see the list of available models with:++```bash+\models available # REPL+plandex models available # CLI+```++## Model Settings++You can see the model settings for the current plan with:++```bash+\models # REPL+plandex models # CLI+```++And you can see the default model settings for new plans with:++```bash+\models default # REPL+plandex models default # CLI+```++You can change the model settings for the current plan with:++```bash+\set-model # REPL+plandex set-model # CLI+```++And you can set the default model settings for new plans with:++```bash+\set-model default # REPL+plandex set-model default # CLI+```++[More details on model settings](./model-settings.md)++## Providers++Plandex offers flexibility on the providers you can use to serve models.++If you use [Plandex Cloud](../hosting/cloud.md) in **Integrated Models Mode**, you can use Plandex credits to pay for AI models. In that case, you won't need to worry about providers, provider accounts, or API keys.++If instead you use **BYO API Key Mode** with Plandex Cloud, or if you [self-host](../hosting/self-hosting/local-mode-quickstart.md) Plandex, you'll need to set API keys (or other credentials) for the providers you want to use. Multiple built-in providers are supported.++If you're self-hosting, you can also configure custom providers—they just need to be OpenAI-compatible.++[More details on providers](./model-providers.md)++## Custom Models, Providers, and Model Packs++You can configure custom models, providers, and model packs with a dev-friendly JSON config file:++```bash+\models custom # REPL+plandex models custom # CLI+```++[More details on custom models, providers, and model packs](./custom-models.md)++## Model Performance++While you can use Plandex with many different providers and models as described above, Plandex's prompts have mainly been written and tested against the built-in models and model packs, so you should expect them to give the best results.++## Local Models++Plandex supports local models via [Ollama](https://ollama.com/). For more details, see the [Ollama Quickstart](./ollama.md).++### Local Models Disclaimer++While local models are supported via Ollama, small models that can be run locally often aren't strong enough to produce usable results for the [heavy-lifting roles](./roles.md) like `planner`, `architect`, `coder`, and `builder`. The prompts for these roles require strong instruction following that can be hard to achieve with small models.++The strongest open source models _are_ capable enough for decent results, but these models are quite large for running locally without a very powerful system. This isn't meant to discourage experimentation with local models, but to set expectations for what is realistically achievable.++To help bridge the gap as local models continue to improve their capabilities, a built-in `ollama-adaptive` model pack is available. This model pack uses local Ollama models for less demanding roles, plus larger remote models for heavy-lifting.+++