Models

Select models shown below - for a complete model list with current pricing, see platform.letta.com/models.

Model	Provider	Max Context	Vision Support?
Claude Opus 4.6	Anthropic	200k	✓
Claude Opus 4.5	Anthropic	200k	✓
Claude Sonnet 4.5	Anthropic	200k	✓
Claude Haiku 4.5	Anthropic	200k	✓
GPT-5.2-Codex	OpenAI	272k	✓
GPT-5.2	OpenAI	272k	✓
Gemini 3 Pro	Google	1M	✓
Gemini 3 Flash	Google	1M	✓
GLM 4.7	Z.ai	200k	-
MiniMax M2.1	MiniMax	180k	-
Kimi K2.5	Moonshot	262k	✓

Bring your own keys (BYOK)

Connect your own LLM API keys to use additional models from providers like OpenRouter, or to connect external coding plans. When using BYOK instead of hosted models through the Letta API, you are billed directly by the LLM API provider.

To configure BYOK in Letta Code, use the /connect command (see full docs). Providers can also be configured via the Letta Platform models page.

---

Configuring models in the Letta API

In Letta Code, use /model to change models.

Model handles

Models are specified using provider/model-name formatted handles:

Python

agent = client.agents.create(
    model="anthropic/claude-sonnet-4-5-20250929",
)

TypeScript

const agent = await client.agents.create({
  model: "anthropic/claude-sonnet-4-5-20250929",
});

Providers

Provider	Prefix	Example
OpenAI	openai/	openai/gpt-5.2
Anthropic	anthropic/	anthropic/claude-sonnet-4-5-20250929
Google AI	google_ai/	google_ai/gemini-3-flash
Azure	azure/	azure/gpt-4o
AWS Bedrock	bedrock/	bedrock/anthropic.claude-3-5-sonnet
OpenRouter	openrouter/	openrouter/anthropic/claude-3.5-sonnet
Ollama	ollama/	ollama/llama3.2

Model settings

Configure temperature, max tokens, and other settings:

Python

agent = client.agents.create(
    model="openai/gpt-5.2",
    model_settings={
        "provider_type": "openai",
        "temperature": 0.7,
        "max_output_tokens": 4096,
    },
    context_window_limit=128000
)

TypeScript

const agent = await client.agents.create({
  model: "openai/gpt-5.2",
  model_settings: {
    provider_type: "openai",
    temperature: 0.7,
    max_output_tokens: 4096,
  },
  context_window_limit: 128000,
});

Reasoning settings

For OpenAI reasoning models (o1, o3):

agent = client.agents.create(
    model="openai/o3-mini",
    model_settings={
        "provider_type": "openai",
        "reasoning": {
            "reasoning_effort": "medium"  # "low", "medium", or "high"
        }
    }
)

For Claude extended thinking:

agent = client.agents.create(
    model="anthropic/claude-sonnet-4-5-20250929",
    model_settings={
        "provider_type": "anthropic",
        "thinking": {
            "type": "enabled",
            "budget_tokens": 10000
        }
    }
)

Changing models

Update an existing agent’s model:

Python

client.agents.update(
    agent_id=agent.id,
    model="openai/gpt-5.2",
    context_window_limit=64000
)

TypeScript

await client.agents.update(agent.id, {
  model: "openai/gpt-5.2",
  context_window_limit: 64000,
});

Agents keep their memory and tools when you change models.