A tray helper for every AI coding tool: sessions, alerts and usage in one place
About ModelVRAM
Free, in-browser VRAM calculators for running LLMs and image models locally: per-model memory (weights + KV cache), llama.cpp --n-cpu-moe offload planner for MoE models, Qwen-Image 2.1 VRAM calculator, "Can I run X on Y" pages and an open CC BY 4.0 dataset. No signup, no ads.
- Who it is for
- People running LLMs and image models locally (llama.cpp, Ollama, ComfyUI) who need a model, quant and context that fit their GPU.
- Problem it solves
- Model downloads are tens of gigabytes, and it is hard to tell in advance whether weights plus KV cache will fit in VRAM, or how much of a MoE model to offload to the CPU.
- The solution
- Free in-browser calculators estimate per-model memory (weights + KV cache), plan llama.cpp --n-cpu-moe offload for MoE models, estimate Qwen-Image 2.1 VRAM, and answer Can I run X on Y questions before you download anything.
- What makes it unique
- No signup, no ads, runs entirely in the browser, and the underlying model dataset is open under CC BY 4.0 with source code on GitHub.
Features
Per-model VRAM estimates
Weights + KV cache memory for local LLMs across quantizations and context lengths.
Details
- Platforms
- Web
- Pricing
- Free
- Open source
- Built for
People running LLMs and image models locally (llama.cpp, Ollama, ComfyUI) who need a model, quant and context that fit their GPU.
- Building since
2026
Technology
Homepage of modelvram.com
1 / 1Website performance
Independent mobile lab measurements and permanent weekly competition results from IndieTools Speed.
100ExcellentMobile Lighthouse
Last checked
- Load time
- 385 ms
- FCP
- 968 ms
- LCP
- 1.67 s
- CLS
- 0.000
- TTFB
- 3 ms
- TBT
- 0 ms
- Speed index
- 968 ms
Latest Performance Score: 100. Measured: October 7, 2026. Weekly Overall Rank: not finalized. LCP: 1.67 s. Load Time: 385 ms.
- Overall rank
- Not finalized
- Best overall rank
- Not finalized 0 top 10 finishes
- Best category rank
- Not finalized 0 category wins
- Latest change
- 43 ms faster No finalized score yet
Website performance history
Inspect every successful mobile Lighthouse lab measurement by metric and date range.
Range
Metric
Latest score
100 / 100
Higher is better
9 measurements
Use the left and right arrow keys to inspect measurements. Use Home or End to jump to the first or last measurement. Press Escape to dismiss the chart tooltip.
Sep 30, 2026Oct 7, 2026Related products
Others in the same categories.
AI-powered comprehensive platform for effective, self-paced English learning, combining structured learning, interactive practice, rich content and AI
AI-maintained datasets from websites and apps
Stop screening. Start shortlisting.
Nano Banana 2.1 AI image generation and reference editing in one workspace.
Put AI to Work in Your Firm, Before Your Competitors Do