swarmauri_llm_mistral 0.11.0.dev1

Provider-specific Swarmauri import package for Mistral chat, tool-calling, streaming, async, and batch LLM workflows.

Swarmauri Mistral LLM

swarmauri_llm_mistral provides provider-specific Swarmauri imports for the Mistral API. The package exposes MistralModel for chat completion workflows and MistralToolModel for tool-calling workflows that execute toolkit functions and optionally follow up with a second model turn.

The runtime targets Mistral's hosted API at https://api.mistral.ai/v1/, supports the current allowlisted Mistral, Magistral, Codestral, Devstral, Pixtral, OCR, embedding, and moderation model families tracked in the repo, and preserves Swarmauri Conversation and toolkit semantics.

Why Use This Package?

• Keep Mistral-specific imports explicit in Swarmauri applications.
• Use a chat model adapter and a separate tool-calling adapter that match Swarmauri’s llms and tool_llms package boundaries.
• Support sync, async, streaming, and batch generation workflows against the Mistral API.
• Execute tools through MistralToolModel without hand-writing provider-specific function-calling payloads.

FAQ

What does swarmauri_llm_mistral install?

It installs MistralModel under swarmauri.llms and MistralToolModel under swarmauri.tool_llms.

Which Mistral capabilities are wrapped today?

MistralModel wraps chat completion workflows, including streaming and optional JSON response formatting. MistralToolModel wraps tool-calling workflows that serialize Swarmauri tools into Mistral-compatible schemas, execute tool calls, and optionally request a final assistant answer in a second turn.

Does it support usage metadata?

MistralModel can attach Swarmauri UsageData when usage information is returned and include_usage is enabled.

Does MistralToolModel support streaming?

Yes. The tool model performs the tool-call round first, appends tool results to the conversation, then streams the final assistant response when streaming is requested.

Which model families are covered?

The current allowlists include Mistral chat models plus related families such as Magistral, Codestral, Devstral, Pixtral, OCR, embed, and moderation surfaces tracked in the runtime code.

Where should I verify current model and pricing details?

Use the provider documentation links in docs/LLM_PROVIDER_MODEL_PRICING_LINKS.md, especially the Mistral models, function-calling, and pricing pages, before publishing model availability or pricing claims.

Features

• MistralModel for sync, async, streaming, and batch chat completion workflows.
• Optional JSON response mode and safe prompt support on chat requests.
• MistralToolModel for tool-calling and multiturn tool execution workflows.
• Current repo-tracked allowlists for Mistral, Magistral, Codestral, Devstral, Pixtral, OCR, embed, and moderation models.
• Compatibility with Python 3.10, 3.11, 3.12, 3.13, and 3.14.

Installation

uv add swarmauri_llm_mistral

pip install swarmauri_llm_mistral

Usage

Set MISTRAL_API_KEY in your environment before creating the model.

Chat Completion

import os

from swarmauri_llm_mistral import MistralModel
from swarmauri_standard.conversations.Conversation import Conversation
from swarmauri_standard.messages.HumanMessage import HumanMessage

conversation = Conversation()
conversation.add_message(HumanMessage(content="Summarize Swarmauri in two sentences."))

model = MistralModel(
    api_key=os.environ["MISTRAL_API_KEY"],
    name="mistral-medium-2508",
)
result = model.predict(conversation=conversation, max_tokens=200)

print(result.get_last().content)

Streaming

import os

from swarmauri_llm_mistral import MistralModel
from swarmauri_standard.conversations.Conversation import Conversation
from swarmauri_standard.messages.HumanMessage import HumanMessage

conversation = Conversation()
conversation.add_message(HumanMessage(content="Write a short haiku about AI tooling."))

model = MistralModel(api_key=os.environ["MISTRAL_API_KEY"])

for token in model.stream(conversation=conversation):
    print(token, end="", flush=True)

Tool Calling

import os

from swarmauri_llm_mistral import MistralToolModel
from swarmauri_standard.conversations.Conversation import Conversation
from swarmauri_standard.messages.HumanMessage import HumanMessage
from swarmauri_standard.toolkits.Toolkit import Toolkit
from swarmauri_standard.tools.AdditionTool import AdditionTool

conversation = Conversation()
conversation.add_message(HumanMessage(content="What is 19 + 23? Use the tool."))

toolkit = Toolkit(tools={"add": AdditionTool()})

model = MistralToolModel(
    api_key=os.environ["MISTRAL_API_KEY"],
    name="mistral-medium-2508",
)
result = model.predict(
    conversation=conversation,
    toolkit=toolkit,
    tool_choice="auto",
)

print(result.get_last().content)

Examples

• Use MistralModel for standard chat completion workflows when you want a direct Mistral provider package import.
• Use MistralToolModel when the application needs function calling and tool execution through a Swarmauri toolkit.
• Use enable_json=True on MistralModel when downstream code expects structured JSON output.

Related Packages

• swarmauri_llm_openai
• swarmauri_llm_groq
• swarmauri_llm_hyperbolic
• swarmauri_llm_deepinfra
• swarmauri_llm_perplexity
• swarmauri_llm_llamacpp

Foundational Swarmauri Packages

• swarmauri
• swarmauri_core
• swarmauri_base
• swarmauri_standard

More Documentation

• LLM provider model and pricing links
• Mistral models
• Mistral function calling

Best Practices

• Keep MISTRAL_API_KEY in environment variables or a secret manager.
• Validate model availability against Mistral’s current catalog before hard-coding a specific family into production workflows.
• Use MistralToolModel only when a toolkit-backed tool loop is actually needed; plain chat is simpler with MistralModel.
• Tune temperature, top_p, max_tokens, and safe_prompt to match your product constraints.

License

Apache-2.0