> ## Documentation Index
> Fetch the complete documentation index at: https://docs.bricks.tools/llms.txt
> Use this file to discover all available pages before exploring further.
# Supported models
> AI models available in CTOR
CTOR supports multiple AI model providers. Configure API keys in [settings](/ctor/reference/settings) to enable a provider.
The **Image input** column shows native image support. A model marked **—** can still work with images if you set a [fallback vision model](/ctor/reference/settings#fallback-vision-model) — CTOR describes each image as text so the model can use it.
## Anthropic
| Model | Context window | Reasoning | Extended thinking | Image input |
| ------------------------- | -------------- | --------- | ----------------- | ----------- |
| claude-fable-5 | 1M | Yes | Max | Yes |
| claude-sonnet-5 | 1M | Yes | Max | Yes |
| claude-sonnet-4.6 | 200K | Yes | — | Yes |
| claude-opus-4.8 | 200K | Yes | Max | Yes |
| claude-opus-4.7 | 200K | Yes | Max | Yes |
| claude-opus-4.6 | 200K | Yes | Max | Yes |
| claude-haiku-4-5-20251001 | 200K | Yes | — | Yes |
Anthropic supports two authentication modes: an API key (pay-per-token billing) or **Login with Claude** OAuth (credit-based usage against a Claude Pro or Max subscription). Configure either in **Settings** > **Providers** > **Anthropic**. See [settings](/ctor/reference/settings#anthropic) for details.
## OpenAI
| Model | Context window | Reasoning | Extended thinking | Image input |
| ------------------ | -------------- | --------- | ----------------- | ----------- |
| gpt-5.5 | 272K | Yes | Max | Yes |
| gpt-5.4 | 272K | Yes | Max | Yes |
| gpt-5.3-codex | 400K | Yes | Max | Yes |
| gpt-5.2-codex | 400K | Yes | Max | Yes |
| gpt-5.2 | 400K | Yes | Max | Yes |
| gpt-5.1-codex-mini | 400K | Yes | — | Yes |
| gpt-5.4-nano | 400K | Yes | Max | Yes |
## OpenAI Codex
| Model | Context window | Reasoning | Extended thinking | Image input |
| ------------------- | -------------- | --------- | ----------------- | ----------- |
| gpt-5.5 | 272K | Yes | Max | Yes |
| gpt-5.4 | 272K | Yes | Max | Yes |
| gpt-5.4-mini | 272K | Yes | Max | Yes |
| gpt-5.3-codex-spark | 128K | Yes | Max | — |
`gpt-5.3-codex-spark` requires a ChatGPT Pro plan. It appears in the model selector only when the signed-in OpenAI Codex account is on the Pro plan.
gpt-5.5 and gpt-5.4 support [speed mode](#speed-mode) — a Normal/Fast toggle appears in the input bar when either model is selected.
## Google
| Model | Context window | Reasoning | Extended thinking | Image input |
| ---------------------- | -------------- | --------- | ----------------- | ----------- |
| gemini-3.5-flash | 1M | Yes | — | Yes |
| gemini-3.1-pro-preview | 1M | Yes | — | Yes |
| gemini-3-pro-preview | 1M | Yes | — | Yes |
| gemini-3-flash-preview | 1M | Yes | — | Yes |
| gemma-4-26b-a4b-it | 256K | Yes | — | Yes |
| gemma-4-31b-it | 256K | Yes | — | Yes |
## Google Agent Platform (Vertex AI)
| Model | Context window | Reasoning | Extended thinking | Image input |
| ---------------------- | -------------- | --------- | ----------------- | ----------- |
| gemini-3.5-flash | 1M | Yes | — | Yes |
| gemini-3.1-pro-preview | 1M | Yes | — | Yes |
| gemini-3-flash-preview | 1M | Yes | — | Yes |
| claude-opus-4.7 | 200K | Yes | Max | Yes |
| claude-opus-4.6 | 200K | Yes | Max | Yes |
| claude-sonnet-4.6 | 200K | Yes | — | Yes |
| claude-haiku-4.5 | 200K | Yes | — | Yes |
Google rebranded Vertex AI to Agent Platform; CTOR shows both names so the provider is findable by either label. Gemini models go through the Generative AI API on Vertex; Claude models use Anthropic's Vertex partner endpoint. Authentication uses Application Default Credentials (ADC), a service account key file, or an API key (Gemini only — Claude partner models reject API keys). See [settings](/ctor/reference/settings#google-agent-platform-vertex-ai) for setup details.
## GitHub Copilot
| Model | Context window | Reasoning | Extended thinking | Image input |
| ---------------------- | -------------- | --------- | ----------------- | ----------- |
| claude-sonnet-5 | 1M | Yes | Max | Yes |
| claude-sonnet-4.6 | 200K | Yes | — | Yes |
| claude-haiku-4.5 | 144K | Yes | — | Yes |
| gpt-5.4 | 400K | Yes | Max | Yes |
| gpt-5.4-mini | 400K | Yes | Max | Yes |
| gpt-5.3-codex | 400K | Yes | Max | Yes |
| gpt-5.2 | 264K | Yes | Max | Yes |
| gpt-5.2-codex | 400K | Yes | Max | Yes |
| gemini-3.5-flash | 128K | Yes | — | Yes |
| gemini-3.1-pro-preview | 128K | Yes | — | Yes |
| gemini-3-flash-preview | 128K | Yes | — | Yes |
| kimi-k2.7-code | 256K | Yes | — | Yes |
GitHub Copilot uses OAuth for authentication. Configure it in **Settings** > **Providers** > **GitHub Copilot**.
OpenAI Codex uses OAuth for authentication (**Login with OpenAI**). Configure it in **Settings** > **Providers** > **OpenAI Codex**.
## OpenCode Zen
| Model | Context window | Reasoning | Extended thinking | Image input |
| ----------------- | -------------- | --------- | ----------------- | ----------- |
| claude-fable-5 | 1M | Yes | Max | Yes |
| claude-sonnet-5 | 1M | Yes | Max | Yes |
| claude-sonnet-4.6 | 200K | Yes | — | Yes |
| claude-opus-4.7 | 200K | Yes | Max | Yes |
| claude-opus-4.6 | 200K | Yes | Max | Yes |
| claude-haiku-4.5 | 200K | Yes | — | Yes |
| gpt-5.4 | 272K | Yes | Max | Yes |
| gpt-5.4-mini | 400K | Yes | Max | Yes |
| gpt-5.3-codex | 400K | Yes | Max | Yes |
| gpt-5.2-codex | 400K | Yes | Max | Yes |
| gpt-5.2 | 400K | Yes | Max | Yes |
| gemini-3.1-pro | 1M | Yes | — | Yes |
| gemini-3-flash | 1M | Yes | — | Yes |
| kimi-k2.7-code | 256K | Yes | — | Yes |
| kimi-k2.6 | 256K | Yes | — | Yes |
| kimi-k2.5 | 256K | Yes | — | Yes |
| glm-5.2 | 1M | Yes | Max | — |
| glm-5.1 | 200K | Yes | — | — |
| glm-5 | 200K | Yes | — | — |
| minimax-m3 | 512K | Yes | — | Yes |
| minimax-m2.5 | 200K | Yes | — | — |
## OpenCode Go
| Model | Context window | Reasoning | Extended thinking | Image input |
| ----------------- | -------------- | --------- | ----------------- | ----------- |
| kimi-k2.7-code | 256K | Yes | — | Yes |
| kimi-k2.6 | 256K | Yes | — | Yes |
| kimi-k2.5 | 256K | Yes | — | Yes |
| glm-5.2 | 1M | Yes | Max | — |
| glm-5.1 | 200K | Yes | — | — |
| minimax-m3 | 512K | Yes | — | Yes |
| minimax-m2.7 | 200K | Yes | — | — |
| qwen3.7-max | 1M | Yes | — | — |
| qwen3.7-plus | 256K | Yes | — | Yes |
| qwen3.6-plus | 256K | Yes | — | Yes |
| mimo-v2.5-pro | 1M | Yes | — | — |
| mimo-v2.5 | 1M | Yes | — | Yes |
| deepseek-v4-pro | 1M | Yes | Max | — |
| deepseek-v4-flash | 1M | Yes | Max | — |
## OpenAI Compatible endpoints
Add any number of OpenAI-compatible API endpoints, each with its own name, base URL, API key, and model list. Manage endpoints in **Settings** > **OpenAI Compatible Endpoints**.
Each endpoint is configured with:
| Field | Description |
| ----------------- | ------------------------------------------------------------------------------------------------------------------------------------------------ |
| **Endpoint Name** | A display name (e.g., "Ollama", "vLLM", "LM Studio") |
| **Base URL** | The API endpoint (e.g., `http://localhost:11434/v1`) |
| **API Mode** | **Chat Completions** (default) or **Responses** — selects which OpenAI API the endpoint uses. Choose **Responses** for reasoning-capable models. |
| **API Key** | Optional, depending on the provider |
| **Models** | One or more model identifiers available at this endpoint |
Models from configured endpoints appear in the model selector grouped under the endpoint name. This allows you to use local models (Ollama, LM Studio) or any third-party provider that implements the OpenAI API format.
If you previously configured a single OpenAI Compatible provider, it is automatically migrated to the new multi-endpoint format on first launch.
### Per-model options
Each model you add to an endpoint carries its own options:
| Option | Description |
| ------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Vision** | Treat the model as supporting image input. CTOR does not verify endpoint support. |
| **Reasoning** | Treat the model as a reasoning-capable Responses model. CTOR sends `reasoning.effort` and preserves encrypted reasoning across turns. Available only when the endpoint's **API Mode** is **Responses**. |
| **XHigh** | Allow the `xhigh` thinking level for this model. Available only when **Reasoning** is enabled. |
| **Context** | Override the context window in tokens (minimum 32K). Leave blank for the 128K default. |
The reasoning options unlock in sequence: set the endpoint **API Mode** to **Responses** to enable **Reasoning**, then enable **Reasoning** to unlock **XHigh**.
## Thinking levels
Models with reasoning support can use different thinking levels to control the depth of analysis:
| Level | Token budget | Best for |
| -------- | ------------ | ----------------------------------------- |
| **Off** | — | Fast responses, simple tasks |
| **High** | \~16k tokens | Most tasks |
| **Max** | \~32k tokens | Complex reasoning, architecture decisions |
The **Max** level is only available on models marked with "Max" in the extended thinking column above.
## Speed mode
When using OpenAI Codex with gpt-5.5 or gpt-5.4, a speed mode selector (gauge icon) appears to the right of the thinking level selector in the input bar.
| Mode | Behavior |
| ---------- | ---------------------------------------------- |
| **Normal** | Standard processing — default latency and cost |
| **Fast** | Priority tier — lower latency, higher cost |
**Fast** maps to the Codex Responses API `service_tier: 'priority'`. The setting is persisted per session.
The selector is only visible when a model that supports speed mode is active. Switching to a model without speed mode support hides the control; any previously set "Fast" preference is not applied.
## Environment variable fallback
If no API key is configured in the app settings, the agent checks these environment variables:
* `ANTHROPIC_API_KEY` — used for Anthropic models
* `OPENAI_API_KEY` — used for OpenAI models