Qwen3.8-Flash-Next · IQ2_XS
This page aggregates 1 real-world runs of Qwen3.8-Flash-Next (IQ2_XS) on NVIDIA RTX 5070 with Strata, contributed by 1 independent source platforms; metrics are averages of published measurements.
79 tok/s
Decode
Decode speed
2,090 tok/s
Prefill
Prefill speed
1 s
TTFT
Time to first token
7.5 GB
VRAM
VRAM usage
1
Measured runs
1
Independent sources
Other
Source platforms
Today
Last verified
Performance
- Strata · IQ2_XS (current)Decode 79 · Prefill 2090 ·
Core figures
- Decode (avg)
- 79 tok/s
- Prefill (avg)
- 2,090 tok/s
- TTFT (avg)
- 1 s
- VRAM (avg)
- 7.5 GB
- MTP acceptance rate
- —
- TTFB
- — GB
- Power draw
- 220 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Qwen3.8-Flash-Next
- Quantization
- IQ2_XS
- Framework
- Strata
- Version
- 0.5.2
- Context length
- 4096 tokens
- Batch size
- 1
Hardware
Nominal and measured figures are shown side by side; whether it runs is the reader's call
- GPU
- NVIDIA RTX 5070
- Nominal VRAM
- 12 GB
- Measured VRAM (avg)
- 7.5 GB
- OS
- Ubuntu 24.04.4 LTS
- Power draw
- 220 W
Get started
Original model resources
Base model resources may not match this quantization. Use the verified deployment weights above when available.
Sources & evidence
1 measured records in total, each traceable to its original source
0.5.2 · Ubuntu 24.04.4 LTS · 4096 ctx
79 tok/s
Decode
2,090 tok/s
Prefill
1 s
TTFT
7.5 GB
VRAM
—
MTP
220 W
Power draw
RTX 5070 12GB Blackwell + Strata 0.5.2 + IQ2_XS：short decode 79 tok/s，128K ctx 63 tok/s，prompt 2090 tok/s，TTFT 1.0s。Strata 官方 RTX 5070 12GB IQ2_XS 基准，384 GB/s 内存带宽 + MTP 推测解码。