Image-Text-to-Text
MLX
Safetensors
qwen3_5
clairic
cloudflare
systemone
structured-output
information-extraction
multimodal
custom-code
conversational
4-bit precision
Instructions to use mlx-community/clairic-flash-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use mlx-community/clairic-flash-4bit with MLX:
# Make sure mlx-vlm is installed # pip install --upgrade mlx-vlm from mlx_vlm import load, generate from mlx_vlm.prompt_utils import apply_chat_template from mlx_vlm.utils import load_config # Load the model model, processor = load("mlx-community/clairic-flash-4bit") config = load_config("mlx-community/clairic-flash-4bit") # Prepare input image = ["http://images.cocodataset.org/val2017/000000039769.jpg"] prompt = "Describe this image." # Apply chat template formatted_prompt = apply_chat_template( processor, config, prompt, num_images=1 ) # Generate output output = generate(model, processor, formatted_prompt, image) print(output) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Pi
How to use mlx-community/clairic-flash-4bit with Pi:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "mlx-community/clairic-flash-4bit"
Configure the model in Pi
# Install Pi: npm install -g @earendil-works/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "mlx-lm": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "mlx-community/clairic-flash-4bit" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Hermes Agent
How to use mlx-community/clairic-flash-4bit with Hermes Agent:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "mlx-community/clairic-flash-4bit"
Configure Hermes
# Install Hermes: curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash hermes setup # Point Hermes at the local server: hermes config set model.provider custom hermes config set model.base_url http://127.0.0.1:8080/v1 hermes config set model.default mlx-community/clairic-flash-4bit
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use mlx-community/clairic-flash-4bit with OpenClaw:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "mlx-community/clairic-flash-4bit"
Configure OpenClaw
# Install OpenClaw: npm install -g openclaw@latest # Register the local server and set it as the default model: openclaw onboard --non-interactive --mode local \ --auth-choice custom-api-key \ --custom-base-url http://127.0.0.1:8080/v1 \ --custom-model-id "mlx-community/clairic-flash-4bit" \ --custom-provider-id mlx-lm \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
Download clairic_runtime/contracts.py from mlx-community/clairic-flash-4bit: direct link, hf CLI and curl.
- Browser
- Download file 6.11 kB
-
https://huggingface.co/mlx-community/clairic-flash-4bit/resolve/main/clairic_runtime/contracts.py
- Command line
-
hf download hf://mlx-community/clairic-flash-4bit/clairic_runtime/contracts.py
-
curl -L -o contracts.py https://huggingface.co/mlx-community/clairic-flash-4bit/resolve/main/clairic_runtime/contracts.py
6.11 kB
| from __future__ import annotations | |
| from typing import Any | |
| from jsonschema import Draft202012Validator, SchemaError | |
| COMMON = {"type", "instructions", "nullable"} | |
| NUMERIC = {"integer", "decimals", "unit"} | |
| GRAMMAR_NUMBER = {"minimum", "maximum", "exclusiveMinimum", "exclusiveMaximum", "enum"} | |
| OPTIONS = { | |
| "text": {"knowledge", "minLength", "maxLength", "pattern", "enum"}, | |
| "number": NUMERIC | GRAMMAR_NUMBER | {"knowledge", "source_only"}, | |
| "compute": NUMERIC | {"null_threshold"}, | |
| "date": {"null_threshold", "source_only"}, | |
| "time": {"null_threshold", "source_only"}, | |
| "datetime": {"null_threshold", "source_only"}, | |
| "duration": NUMERIC | {"null_threshold"}, | |
| "boolean": set(), | |
| "enum": {"enum"}, | |
| "object": {"properties"}, | |
| "list": {"items", "max_items", "knowledge", "source_only"} | NUMERIC | {"enum"}, | |
| "estimate": {"min", "max", "scale", "step", "unit"}, | |
| "forecast": {"min", "max", "scale", "step", "unit", "horizon", "interval"}, | |
| } | |
| def item_spec(spec: dict[str, Any]) -> dict[str, Any]: | |
| item = spec["items"] | |
| return dict(item) if isinstance(item, dict) else {"type": item} | |
| def validate_public_fields(fields: dict[str, dict[str, Any]]) -> None: | |
| _validate_public_fields(fields, set()) | |
| def _validate_public_fields( | |
| fields: dict[str, dict[str, Any]], checked_scalar_schemas: set[tuple[str, ...]] | |
| ) -> None: | |
| for name, spec in fields.items(): | |
| kind = spec.get("type") if isinstance(spec, dict) else None | |
| if not isinstance(kind, str) or kind not in OPTIONS: | |
| raise ValueError(f"field {name}: unknown field type") | |
| unknown = spec.keys() - COMMON - OPTIONS[kind] | |
| if unknown: | |
| raise ValueError(f"field {name}: unknown options {sorted(unknown)}") | |
| if "instructions" in spec and (not isinstance(spec["instructions"], str)): | |
| raise ValueError(f"field {name}: instructions must be a string") | |
| if "nullable" in spec and (not isinstance(spec["nullable"], bool)): | |
| raise ValueError(f"field {name}: nullable must be true or false") | |
| if "source_only" in spec: | |
| if not isinstance(spec["source_only"], bool): | |
| raise ValueError(f"field {name}: source_only must be true or false") | |
| if kind == "list" and spec.get("items") not in ("date", "number"): | |
| raise ValueError( | |
| f"field {name}: source_only lists require date or number items" | |
| ) | |
| if spec["source_only"] and "null_threshold" in spec: | |
| raise ValueError( | |
| f"field {name}: source_only does not support typed null_threshold" | |
| ) | |
| if spec["source_only"] and spec.get("knowledge"): | |
| raise ValueError( | |
| f"field {name}: source_only cannot permit knowledge answers" | |
| ) | |
| if "enum" in spec and (not isinstance(spec["enum"], list) or not spec["enum"]): | |
| raise ValueError(f"field {name}: enum must be a nonempty array") | |
| if kind == "enum" and "enum" not in spec: | |
| raise ValueError(f"field {name}: enum values are required") | |
| if kind == "object": | |
| properties = spec.get("properties") | |
| if not isinstance(properties, dict) or not properties: | |
| raise ValueError(f"field {name}: object properties are required") | |
| if any( | |
| ( | |
| not isinstance(child, dict) | |
| or child.get("type") not in ("text", "number", "boolean", "enum") | |
| for child in properties.values() | |
| ) | |
| ): | |
| raise ValueError( | |
| f"field {name}: row properties must be text, number, boolean or enum" | |
| ) | |
| if any((child.get("source_only") for child in properties.values())): | |
| raise ValueError( | |
| f"field {name}: source_only is supported on scalar numbers and number lists, not row properties" | |
| ) | |
| _validate_public_fields(properties, checked_scalar_schemas) | |
| from . import schema as zs | |
| zs.validate_fields(properties) | |
| if kind == "list": | |
| if "items" not in spec: | |
| raise ValueError(f"list {name}: items are required") | |
| child = item_spec(spec) | |
| if isinstance(spec["items"], dict) and child.get("type") != "object": | |
| raise ValueError(f"list {name}: structured items must be objects") | |
| if child.get("type") == "object": | |
| _validate_public_fields({name + "[]": child}, checked_scalar_schemas) | |
| if spec.keys() & (NUMERIC | {"knowledge", "enum"}): | |
| raise ValueError( | |
| f"list {name}: put numeric or knowledge options on row properties" | |
| ) | |
| maximum = spec.get("max_items") | |
| if "max_items" in spec and ( | |
| isinstance(maximum, bool) or not isinstance(maximum, int) or maximum < 0 | |
| ): | |
| raise ValueError( | |
| f"list {name}: max_items must be a nonnegative integer" | |
| ) | |
| if kind in ("text", "number", "boolean", "enum", "object"): | |
| from . import ar | |
| try: | |
| schema = ar.schema(spec) | |
| scalar_key = None | |
| if schema.keys() == {"type"}: | |
| schema_type = schema["type"] | |
| scalar_key = ( | |
| tuple(schema_type) | |
| if isinstance(schema_type, list) | |
| else (schema_type,) | |
| ) | |
| if scalar_key is None or scalar_key not in checked_scalar_schemas: | |
| Draft202012Validator.check_schema(schema) | |
| if scalar_key is not None: | |
| checked_scalar_schemas.add(scalar_key) | |
| except SchemaError as error: | |
| raise ValueError( | |
| f"field {name}: invalid schema constraint: {error.message}" | |
| ) from error | |