Systems
Reading live sources…
Systems
Reading live sources…
Reading live data…
The open models FlexInfer is serving right now, the GPUs they sit on, and what each deployment measured. Each panel is read when the page loads and says where its numbers came from.
FlexDeck’s metrics summary could not be read, so there are no plane figures to show. The model list and benchmarks below have their own sources.
FlexDeck’s model status could not be read, so there is no model list to show right now.
One row per run of the FlexInfer benchmarker against a deployed model; the newest record is from . These are operational baselines on this hardware, not a cross-vendor leaderboard.
Decode throughput the benchmarker measured for each run, highest first. 2 runs report other units (emb/s, img/min) and are not ranked here.
| Model | Backend | GPU | Throughput | Tokens | Seconds | Samples | Date |
|---|---|---|---|---|---|---|---|
| qwen3-1p7b-tools-radeonvii | llama.cpp | Radeon VII | 23.8 tok/s | 320 | 13.5 | 10 | Sep 2, 2026 |
| gemma4-26b-a4b-gptq-5930k | vllm | Radeon RX 7900 XTX | 142.1 tok/s | 3,136 | 22.1 | 49 | Aug 30, 2026 |
| gemma4-26b-a4b-gptq | vllm | Radeon RX 7900 XTX | 139.3 tok/s | 3,008 | 21.6 | 47 | Aug 30, 2026 |
| BAAI/bge-large-en-v1.5 | tei | CPU | 203.25 emb/s | — | — | — | Jan 21, 2026 |
| Qwen3-14B-q4f16_1-MLC | mlc-llm | AMD · gfx1100 | 82.4 tok/s | — | — | — | Jan 21, 2026 |
| stabilityai/sdxl-turbo | diffusers | AMD · gfx1100 | 61.6 img/min | — | — | — | Jan 21, 2026 |
| qwen35-35b-clean-gptq-64k-graph-candidate (candidate) | vllm | Not recorded | 79.4 tok/s | 672 | 8.5 | 21 | Jul 14, 2026 |
This page renders the same documents the public API serves, after the anonymizers in lib/public-api/anonymize.ts have run on the server.
The nodes under these models are on the cluster page.
© 2026 Cody Blevins.