Provider models¶

In addition to the regular pydantic-ai models, LLMling-agent supports all model types from llmling-models through YAML configuration. Each model is identified by its type field. These models often are some kind of "meta-models", allowing model selection patterns as well as human-in-the-loop interactions.

Basic Configuration¶

agents:
  my_agent:
    model:
      type: string           # Basic string model identifier
      identifier: gpt-5      # Model name

Available Model Types¶

See Config section to see the available types.

Model Settings¶

You can set common model settings to fine-tune the LLM behavior:

agents.yml

# yaml-language-server: $schema=https://raw.githubusercontent.com/phil65/llmling-agent/refs/heads/main/schema/config-schema.json
agents:
  tuned_agent:
    model:
      identifier: openai:gpt-10-turbo-ultra-mega
      max_tokens: 2000 # Maximum tokens to generate
      temperature: 0.7 # Randomness (0.0 - 2.0)
      top_p: 0.9 # Nucleus sampling threshold
      timeout: 30.0 # Request timeout in seconds
      parallel_tool_calls: true # Allow parallel tool execution
      seed: 42 # Random seed for reproducibility
      presence_penalty: 0.5 # (-2.0 to 2.0) Penalize token reuse
      frequency_penalty: 0.3 # (-2.0 to 2.0) Penalize token frequency
      logit_bias: # Modify token likelihood
        "1234": 100 # Increase likelihood
        "5678": -100 # Decrease likelihood

  ### Example with Provider and Model Settings

  advanced_agent:
    name: "Advanced GPT-5"
    end_strategy: early
    model:
      type: string
      identifier: openai:gpt-5
      temperature: 0.8
      max_tokens: 1000
      presence_penalty: 0.2
      timeout: 60.0

  cautious_agent:
    name: "Careful Claude"
    model:
      identifier: anthropic:claude-sonnet-4-0
      temperature: 0.3 # More deterministic
      max_tokens: 2000
      timeout: 120.0 # Longer timeout

All settings are optional and providers will use their defaults if not specified.

Setting pydantic-ai models by identifier¶

LLMling-agent also extends pydantic-ai functionality by allowing to define more models via simple string identifiers. These providers are

OpenRouter (openrouter:provider/model-name, requires OPENROUTER_API_KEY env var)
Grok (X) (grok:grok-2-1212, requires X_AI_API_KEY env var)
DeepSeek (deepsek:deepsek-chat, requires DEEPSEEK_API_KEY env var)

For detailed model documentation and features, see the llmling-models repository.