GLM 5.3 Flash NVFP4
Runs as your own dedicated GPU deployment, ready to launch from the wizard. Data stays in Europe.
$ curl https://hostyourai.com/api/v1/chat/completions \
-H "Authorization: Bearer hyai-..." \
-H "Content-Type: application/json" \
-d '{
"model": "nvidia/GLM-5.3-Flash-NVFP4",
"messages": [{"role": "user", "content": "Hello!"}]
}'
Call it now
Drop-in replacement for OpenAI: change only the base URL and API key. The Anthropic format (/v1/messages) is supported too.
text+image->text · nvidia · sovereign EU
Runs on EU infrastructure operated by European companies, from your request to the GPU. See the full chain
This model is ready for commercial or non-commercial use. <br
- Router
- Router · on request
- Dedicated
- Dedicated · available
Specifications
| Parameters | 321B |
|---|---|
| Context window | 1,048,576 tokens |
| Minimum VRAM | 246 GB |
| Architecture | Glm5NextForConditionalGeneration (vLLM) |
| License | mit |
| Modality | text+image->text |
| Released | September 2026 |
| Publisher | nvidia ↗ |
Pricing
Shared router · per token
Dedicated GPU · per hour
Shared EU router, you pay only per token. Dedicated GPU deployments are billed per minute, see pricing.
Bench index
Bench is our quality index for every model, 0 to 100. It rests on public measurements (Epoch AI with about 40 benchmarks, LiveBench, LMArena and more) and is refreshed every morning. Where nobody measures, it shows an estimate with a range.
Translated from its LiveBench result onto the same scale.
Range 67.3 to 81.4
Sources: Epoch AI ↗ · LiveBench ↗ · LMArena ↗ · Refreshed daily, last on 09-11-2026
LiveBench · 71.6 · Rank 55 of 70
LiveBench · All tasks
| theory of mind | 73.1 |
|---|---|
| zebra puzzle | 79.5 |
| spatial | 96.0 |
| logic with navigation | 62.0 |
| code generation | 77.5 |
|---|---|
| code completion | 80.4 |
| javascript | 63.6 |
|---|---|
| typescript | 46.7 |
| python | 60.0 |
| AMPS Hard | 97.0 |
|---|---|
| integrals with game | 45.0 |
| math comp | 93.1 |
| olympiad | 89.8 |
| consecutive events | 84.3 |
|---|---|
| tablejoin | 48.8 |
| tablereformat | 96.1 |
| connections | 93.7 |
|---|---|
| plot unscrambling | 60.3 |
| typos | 78.0 |
| paraphrase | 48.0 |
|---|---|
| simplify | 47.0 |
| story generation | 59.9 |
| summarize | 56.5 |
Frequently asked questions
Can I run GLM 5.3 Flash NVFP4 in the EU?
Yes. HostYourAI runs GLM 5.3 Flash NVFP4 on GPUs in European datacenters via vLLM. Prompts and outputs never leave the EU and there is no US cloud provider in the chain.
Is hosting GLM 5.3 Flash NVFP4 GDPR-compliant?
Yes. All processing happens inside the EU, a Data Processing Agreement (DPA) is available and the subprocessor list is public. Open-source weights also mean: no training on your data.
How much does GLM 5.3 Flash NVFP4 cost?
GLM 5.3 Flash NVFP4 is available on request: through the shared EU router after a quick chat, or as your own dedicated GPU deployment billed per minute. Tell us your use case and we will set it up for you.
Is the API OpenAI-compatible?
Yes. You use the standard OpenAI SDKs with a custom base URL (https://hostyourai.com/api/v1). The Anthropic Messages API is supported as a drop-in as well.
More models from NVIDIA
One platform, four layers
Start with one layer and add the others when the work asks for it. They share one key, one balance and one Activity view.
Try GLM 5.3 Flash NVFP4 for free
Creating an account takes a minute. Test GLM 5.3 Flash NVFP4 straight away in the playground.
