Ministral 3 8B Instruct 2512 is an open-source chat model from Mistral that takes in text and images and outputs text. It is designed for instruction-based tasks, including calling tools for code execution and function calling.
Input
Output
Context
262K
Max Output
262K
Parameters
8.9B
Input Modalities
Output Modalities
Some values come from web research and may be incorrect. Report incorrect data
None the catalog can show from stated facts or measured scores.
Download 35.7 GB · Runs on vLLM v0.32.0 · 11 GGUF builds
| Precision | Weights | One request at 33K tokens | Smallest GPU that fits |
|---|---|---|---|
| FP16 / BF16 | 16.6 GB | 21.0 GB | 1x NVIDIA GeForce RTX 3090 (24 GB) |
| INT8 | 8.31 GB | 12.3 GB | 1x NVIDIA T4 (16 GB) |
| INT4 | 4.15 GB | 7.94 GB | 1x NVIDIA T4 (16 GB) |
Computed by Inferbase from the 8,918,026,240 parameters in the creator's Hugging Face repository; one request adds its KV cache, activations and framework overhead, with the layer layout estimated from the parameter count.
Public resources
Specification sources
As of Nov 17, 2026.
Answered from the facts on this page, with their sources.
Ministral 3 8B Instruct 2512 has a context window of 262,144 tokens and a maximum output of 262,144 tokens (source: Creator or provider API). A provider may serve less; Inferbase routes on what each provider serves.
Yes. Ministral 3 8B Instruct 2512 is an open-weight model under the Apache 2.0 license. You can download and self-host it within the license terms.
Ministral 3 8B Instruct 2512 has 8,918,026,240 parameters (8.9B), as stated in the creator's Hugging Face repository.
Strengths: Open weights under a permissive license (apache-2.0); 262K token context window; Supports 11 languages. Each is computed from the model's stated facts and benchmark results.
Ministral 3 8B Instruct 2512 accepts Text and Image and produces Text. The Capabilities card lists the API features it supports, such as function calling and structured output.
No benchmark source has published a result for Ministral 3 8B Instruct 2512 yet, so it is shown as unmeasured, not estimated.
Not on Inferbase Managed today. Add your own provider key or endpoint that serves Ministral 3 8B Instruct 2512, and Inferbase runs it on your account with the same routing and logs.
The weights take 16.6 GB at FP16 / BF16, 8.3 GB at INT8, 4.2 GB at INT4, computed from the 8,918,026,240 parameters its creator publishes. Each request adds its KV cache and the framework's overhead. The hardware section shows one request and the smallest GPU that fits; the GPU planner sizes it for your context length and traffic.
The model card of Ministral 3 8B Instruct 2512 lists 11 languages: English, French, Spanish, German, Italian, Portuguese, Dutch, Chinese, Japanese, Korean and Arabic. That is what Mistral states; the catalog does not test language quality.
Start on the free plan with 5,000 requests a month, no card required.