llm-benchmarks
githubdrose.io

nextbit Provider Benchmarks


Measured over the last 30 days

Models Tracked
16
Avg Tokens / Second
15.18
Avg Time to First Token (ms)
10004.49
Last Updated
Sep 1, 2026

All nextbit models

16 tracked · fastest first
ModelAvg Toks/SecMinMaxAvg TTF (ms)
UnslopNemo 12B36.1015.0048.001800.00
Gemma 4 26B A4B31.206.2749.002390.00
Gemma 2 27B30.6028.2031.901220.00
Gemma 2 27B30.4020.3043.101300.00
MythoMax 13B27.5015.1039.401140.00
MythoMax 13B26.4014.1041.001350.00
UnslopNemo 12B17.8017.8017.802930.00
ReMM SLERP 13B17.5016.1019.301290.00
Llama 3.3 Euryale 70B13.509.3220.001800.00
Llama 3.3 Euryale 70B12.001.0722.104090.00
DeepSeek V4 Flash 04239.619.619.616430.00
Qwen3 14B8.758.309.205990.00
Gemma 4 26B A4B8.608.608.606410.00
Qwen3 14B7.535.838.457500.00
DeepSeek V4 Pro 04235.735.735.7310700.00
DeepSeek V4 Flash 07313.493.493.4917900.00

Frequently Asked Questions

Which nextbit model is fastest?
Based on recent tests, UnslopNemo 12B shows the highest average throughput among tracked nextbit models.
How many recent measurements feed this dashboard?
This provider summary aggregates 786 individual prompts measured across 684 monitoring runs over the past month.