Flint-40B-A2.5B · IQ3_S
This page aggregates 1 real-world runs of Flint-40B-A2.5B (IQ3_S) on NVIDIA RTX Spark with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.
91.56 tok/s
Decode
Decode speed
457.14 tok/s
Prefill
Prefill speed
0.25 s
TTFT
Time to first token
57.27 GB
VRAM
VRAM usage
1
Measured runs
1
Independent sources
Other
Source platforms
14 days ago
Last verified
Performance
- llama.cpp · IQ3_S (current)Decode 91.56 · Prefill 457.14 ·
- llama.cpp · Q3_K_S Decode 100.54 · Prefill 475.16 ·
- llama.cpp · Q4_K_M Decode 76.22 · Prefill 300.34 ·
- Ollama · IQ4_XS Decode 74.46 · Prefill 328.19 ·
- Ollama · Q2_K Decode 96.77 · Prefill 510.91 ·
- Ollama · Q4_K_M Decode 70.03 · Prefill 271.9 ·
Core figures
- Decode (avg)
- 91.56 tok/s
- Prefill (avg)
- 457.14 tok/s
- TTFT (avg)
- 0.25 s
- VRAM (avg)
- 57.27 GB
- MTP acceptance rate
- —
- TTFB
- 0.73 GB
- Power draw
- 80 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Flint-40B-A2.5B
- Quantization
- IQ3_S
- Framework
- llama.cpp
- Version
- v0.26
- Context length
- 8192 tokens
- Batch size
- 256
- GPU layers
- 20
- Flash Attention
- On
Hardware
Nominal and measured figures are shown side by side; whether it runs is the reader's call
- GPU
- NVIDIA RTX Spark
- Nominal VRAM
- 128 GB
- Measured VRAM (avg)
- 57.27 GB
- OS
- Ubuntu 24.04
- Driver
- 615.78
- CUDA
- 13.0
- Power draw
- 80 W
- Reference price
- $2,399 · updated 2026-10-19
Prices are not live; check the store page
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
v0.26 · Ubuntu 24.04 · CUDA 13.0 · 8192 ctx
91.56 tok/s
Decode
457.14 tok/s
Prefill
0.25 s
TTFT
57.27 GB
VRAM
—
MTP
80 W
Power draw
L2 batch seed 2026-10-22: llama.cpp on hw_id=110 running model_id=92 at IQ3_S