MedGemma is a collection of open-source, instruction-tuned language models from Google. It is a multimodal model that accepts both text and image inputs to generate text outputs.
Input
Output
Context
128K
Max Output
8K
Parameters
4.3B
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 8.60 GB · Runs on vLLM v0.32.0 · 39 GGUF builds
| Precision | Weights | One request at 33K tokens | Smallest GPU that fits |
|---|---|---|---|
| FP16 / BF16 | 8.01 GB | 11.6 GB | 1x NVIDIA T4 (16 GB) |
| INT8 | 4.00 GB | 7.35 GB | 1x NVIDIA T4 (16 GB) |
| INT4 | 2.00 GB | 5.25 GB | 1x NVIDIA T4 (16 GB) |
Computed by Inferbase from the 4,300,079,472 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.
Specification sources
As of Nov 10, 2026.
Answered from the facts on this page, with their sources.
MedGemma 4B IT has a context window of 128,000 tokens and a maximum output of 8,192 tokens (source: Web research). A provider may serve less; Inferbase routes on what each provider serves.
Yes. MedGemma 4B IT is an open-weight model under the Other license. You can download and self-host it within the license terms.
MedGemma 4B IT has 4,300,079,472 parameters (4.3B), as stated in the creator's Hugging Face repository.
MedGemma 4B IT 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 MedGemma 4B IT yet, so it is shown as unmeasured, not estimated.
Not on Inferbase Managed today. Add your own provider key or endpoint that serves MedGemma 4B IT, and Inferbase runs it on your account with the same routing and logs.
The weights take 8.0 GB at FP16 / BF16, 4.0 GB at INT8, 2.0 GB at INT4, computed from the 4,300,079,472 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.
Start on the free plan with 5,000 requests a month, no card required.