Qwen3.5 9B Base
Trinity Nano Preview
1.1T
- Context
- 33Ktokens
- Released
- 2y ago
- Size
- 7Bparameters
Mistral 7B Instruct v0.3 is an instruct fine-tune of the Mistral-7B-v0.3 base model, created by Mistral AI as the v0.3 entry in its Mistral 7B Instruct series. The model card presents it as a demonstration that the base model can be readily fine-tuned for strong performance. It is built for chat-based instruction following and function calling.
How it compares
Size class
Small7BAn open-weight model of 5 to 25 billion parameters, the size that runs on a single consumer GPU.
Open models of a similar size from other labs
Vs. your stack
Add the models you use to see how this one compares.
Sign inAdditional source evidence
- Source model
- mistral/mistral-7b-instruct-v0.3
- Source family
- mistral
- Last updated
- 2024-05-22
- Model context
- 33K
- Model output limit
- 33K
- Open weights
- Yes
- File attachments
- No
- Reasoning
- No
- Tool calling
- Yes
- Temperature control
- Yes
Downloadable weights
Complete source record (JSON)
{
"id": "mistral/mistral-7b-instruct-v0.3",
"name": "Mistral 7B Instruct v0.3",
"description": "Open 7B Mistral instruction model with function calling",
"family": "mistral",
"last_updated": "2024-05-22",
"attachment": false,
"reasoning": false,
"tool_call": true,
"temperature": true,
"modalities": {
"input": [
"text"
],
"output": [
"text"
]
},
"limit": {
"context": 32768,
"output": 32768
},
"release_date": "2024-05-22",
"open_weights": true,
"license": "Apache 2.0",
"weights": [
{
"label": "Hugging Face",
"url": "https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3"
}
]
}Frequently asked questions
Running locally
At Q4 an estimated 5.4 GB, at Q8 9.1 GB, and at FP16 16.6 GB. All VRAM figures are estimates based on the 7B parameter count, weights plus about 15% and 0.5 GB for a short context.
Yes, it fits on all four. A 24 GB GPU fits at FP16 (estimated 16.6 GB), 16 GB and 12 GB GPUs fit at Q8 (estimated 9.1 GB), and an 8 GB GPU fits at Q4 (estimated 5.4 GB).
Yes. The Q4 estimate is 5.4 GB, and under the assumption that macOS lets the GPU use about 75% of unified memory, the smallest common Mac memory size that fits Q4 is an estimated 8 GB.
Privacy and data
Yes, you can self-host it for full control of your data because it has open weights.
Capabilities and specs
The context window is 33K tokens (32,768), and the maximum output is also 33K tokens (32,768).
No, it reads text only. Its input modalities are text.
Yes, tool calling (function calling) is supported.
No, a reasoning or thinking mode is not supported.
Quality and benchmarks
The closest alternatives are OpenELM 1.1B by Apple, OpenELM 270M by Apple, and OLMo 1.7 7B by Ai2. They were matched on model kind and release date, with no comparable price.
OpenELM 1.1BMistral 7B Instruct v0.3 vs OpenELM 1.1BOpenELM 270MMistral 7B Instruct v0.3 vs OpenELM 270MOLMo 1.7 7BMistral 7B Instruct v0.3 vs OLMo 1.7 7B
Licensing and access
It is open weights, not fully open source. The license is Apache 2.0.
Yes. The license is Apache 2.0, a permissive license that allows commercial use.
The weights are available from the Hugging Face repository mistralai/Mistral-7B-Instruct-v0.3.
Basics
Mistral AI made it, and it was released on May 22, 2024.
Yes, it has open weights.
Generated from Vertical Tabs data, updated November 4, 2026.
Discussion
No comments yet
Be the first to share your thoughts.