Skip to main content
View source

Mistral AI

View as Markdown

A RocketRide LLM node that connects Mistral AI's chat models to a pipeline.

What it does

Connects Mistral AI models to your pipeline. Used primarily as an llm invoke connection by agents and other nodes that need an LLM. Can also be used directly via lanes.

Uses the official mistralai Python SDK (both the 1.x and 2.x import layouts are supported) and calls the chat.complete endpoint. Requests are sent with temperature: 0.0 for deterministic responses and no max_tokens cap, so the model decides output length.

Prompts are validated against the model's context window before sending. Token counts are estimated heuristically (about 1.25 tokens per word, adjusted for punctuation and long words); no tokenizer round-trip is made. The context limit comes from the selected profile's modelTotalTokens; if not configured, a built-in per-model default is used, falling back to 32,768 tokens for unknown models.

Failed API calls are retried automatically with exponential backoff (see Retries and error handling), and raw provider errors are rewritten into user-friendly messages.

When you save the node configuration, a minimal validation probe (a one-token "Hi" completion) is sent to verify the API key and model before the pipeline ever runs. Validation failures surface as warnings with the provider's error detail.


Configuration

Lanes

Lane inLane outDescription
questionsanswersSend a question directly, receive a generated answer

Fields

The node is configured through a single profile selector plus per-profile fields:

FieldType / DefaultDescription
profilestring, default mistral-largeWhich Mistral model preset to use (see profiles below)
apikeystringMistral AI API key (required for every profile)
modelSourceshared llm.cloud fieldModel source selector
modelstring (custom profile only)Mistral AI model ID
modelTotalTokensnumber, default 32768 (custom)Maximum context length in tokens

model and modelTotalTokens are only editable in the Custom Model profile; every other profile pins them to the values shown in the profiles table below. An explicitly configured modelTotalTokens always overrides the built-in per-model default.


Profiles

Default profile: Mistral Large 3 (mistral-large).

Flagship

ProfileModelContext
Mistral Large 3 (default)mistral-large-2512262,144
Mistral Medium 3.1mistral-medium-2508131,072
Mistral Small 3.2mistral-small-2506131,072

Reasoning

ProfileModelContext
Magistral Medium 1.2magistral-medium-250940,000
Magistral Small 1.2magistral-small-250940,960

Code

ProfileModelContext
Codestralcodestral-2508256,000
Devstral Medium 1.0devstral-medium-2507128,000
Devstral Small 1.1devstral-small-2507128,000

Edge

ProfileModelContext
Ministral 3 14Bministral-14b-2512262,144
Ministral 3 8Bministral-8b-2512262,144
Ministral 3 3Bministral-3b-2512131,072

Custom: specify any Mistral model ID and token limit directly.

Additional profiles

Beyond the curated set, the node ships presets for dated snapshots, -latest aliases, and specialty models:

ProfileModelContext
codestral-latestcodestral-latest8,191
devstral-2512devstral-2512262,144
devstral-latestdevstral-latest256,000
devstral-medium-latestdevstral-medium-latest256,000
magistral-medium-latestmagistral-medium-latest40,000
magistral-small-latestmagistral-small-latest40,000
ministral-14b-latestministral-14b-latest16,384
ministral-8b-latestministral-8b-latest16,384
ministral-3b-latestministral-3b-latest16,384
mistral-large-2411mistral-large-2411131,072
mistral-large-latestmistral-large-latest32,000
mistral-medium-2505mistral-medium-25058,191
mistral-medium-2604mistral-medium-260416,384
mistral-medium-3mistral-medium-3131,072
mistral-medium-3-5mistral-medium-3-516,384
mistral-medium-c21211-r0-75mistral-medium-c21211-r0-7516,384
mistral-medium-latestmistral-medium-latest131,072
mistral-small-2603mistral-small-2603262,144
mistral-small-latestmistral-small-latest131,072
mistral-tiny-2407mistral-tiny-240716,384
mistral-tiny-latestmistral-tiny-latest16,384
mistral-vibe-cli-fastmistral-vibe-cli-fast16,384
mistral-vibe-cli-latestmistral-vibe-cli-latest16,384
mistral-vibe-cli-with-toolsmistral-vibe-cli-with-tools16,384

Retries and error handling

Transient failures (timeouts, network errors, rate limits, HTTP 5xx) are retried with exponential backoff. Retry budget scales with the model tier:

Model tierRetriesBase delay
large models32.0 s
medium / magistral models21.5 s
All other (small / edge) models21.0 s

Non-retryable errors fail immediately. Whatever the cause, the raw API error is mapped to a user-friendly message: authentication failures, rate limits, quota/billing issues, invalid input, model unavailability, server errors, content-policy violations, timeouts, and network problems each get a specific explanation instead of a stack trace.


Authentication

Provide a Mistral AI API key in the apikey field (created in the Mistral AI console). The key is read at startup and never persisted by the node.

The node detects common key mix-ups at initialization and fails with a specific message:

  • Keys starting with sk- are rejected as OpenAI keys.
  • Keys starting with AI are rejected as Google AI / Gemini keys.

Upstream docs


Schema

FieldTypeDescriptionDefault
mistral.profilestringModel
Mistral AI model selection
"mistral-large"
modelstringModel
Mistral AI model
modelTotalTokensnumberTokens
Maximum context length in tokens

Dependencies

  • mistralai
  • mistral-common[sentencepiece]