Arao AI
Back to search results

Qwen3-Coder-Next · IQ2_M

Start a discussion
NVIDIA RTX 509032GBOllama v0.32L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3-Coder-Next (IQ2_M) on NVIDIA RTX 5090 with Ollama, contributed by 1 independent source platforms; metrics are averages of published measurements.

299.22 tok/s

Decode

Decode speed

1,242.06 tok/s

Prefill

Prefill speed

1.78 s

TTFT

Time to first token

1.05 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

2 days ago

Last verified

Performance

  1. Ollama · IQ2_M (current)Decode 299.22 · Prefill 1242.06 ·
  2. Ollama · Q3_K_M Decode 238.15 · Prefill 870.24 ·
  3. Strata · IQ1_S Decode 320.16 · Prefill 1259.65 ·
  4. vLLM · Q5_K_M Decode — · Prefill — ·
  5. vLLM · Q4_K_M Decode — · Prefill — ·
  6. llama.cpp · Q5_K_M Decode — · Prefill — ·
  7. llama.cpp · Q4_K_M Decode — · Prefill — ·

Core figures

Decode (avg)
299.22 tok/s
Prefill (avg)
1,242.06 tok/s
TTFT (avg)
1.78 s
VRAM (avg)
1.05 GB
MTP acceptance rate
—
TTFB
2.39 GB
Power draw
353.8 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3-Coder-Next
Quantization
IQ2_M
Framework
Ollama
Version
v0.32
Context length
8192 tokens
Batch size
512
GPU layers
99
Flash Attention
On

Hardware

Nominal and measured figures are shown side by side; whether it runs is the reader's call

GPU
NVIDIA RTX 5090
Nominal VRAM
32 GB
Measured VRAM (avg)
1.05 GB
OS
macOS 15.0
Driver
555.85
CUDA
12.8
Power draw
353.8 W

Sources & evidence

1 measured records in total, each traceable to its original source

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-16

    v0.32 · macOS 15.0 · CUDA 12.8 · 8192 ctx

    299.22 tok/s

    Decode

    1,242.06 tok/s

    Prefill

    1.78 s

    TTFT

    1.05 GB

    VRAM

    —

    MTP

    353.8 W

    Power draw

    L2 batch seed 2026-10-16: Ollama on hw_id=13 running model_id=58 at IQ2_M