Model Hub

Echo Fleet — 26 models across 5 nodes

Total Models
26
Online
7
Avg Latency
3.7s
Total VRAM
295.5GB

Speed (<7s)(7)

axe-blade-4b

2.5GB · Q4_K_M
Online
JL2 Forge·16GB
codequick-fix
Avg Response
3.9s
VRAM
2.5GB

axe-phantom-8b

4.9GB · Q4_K_M
Online
JL2 Forge·16GB
codereasoning
Avg Response
5.2s
VRAM
4.9GB

axe-spark-3b

1.8GB · Q4_K_M
Online
JL2 Forge·16GB
chatquick-fix
Avg Response
2.1s
VRAM
1.8GB

axe-razor-7b

4.1GB · Q4_K_M
Offline
JL3 Vigil·16GB
codetool-use
Avg Response
4.7s
VRAM
4.1GB

axe-whisper-3b

1.9GB · Q4_K_M
Offline
JLa Ghost·8GB
chatfast-chat
Avg Response
2.8s
VRAM
1.9GB

axe-pulse-4b

2.3GB · Q4_K_M
Offline
JLa Ghost·8GB
quick-fixtool-use
Avg Response
3.5s
VRAM
2.3GB

axe-echo-7b

4.2GB · Q4_K_M
Offline
JLa Ghost·8GB
chatcreative
Avg Response
5.8s
VRAM
4.2GB

Workhorse (7-15s)(9)

axe-sage-14b

8.4GB · Q4_K_M
Offline
JL1 Nova·64GB
analysisresearch
Avg Response
8.1s
VRAM
8.4GB

axe-oracle-27b

16.2GB · Q4_K_M
Offline
JL1 Nova·64GB
reasoningstrategy
Avg Response
12.3s
VRAM
16.2GB

axe-vision-12b

7.1GB · Q4_K_M
Offline
JL1 Nova·64GB
visionmultimodal
Avg Response
9.5s
VRAM
7.1GB

axe-scribe-20b

12.1GB · Q4_K_M
Offline
JL1 Nova·64GB
creativelong-context
Avg Response
10.8s
VRAM
12.1GB

axe-forge-13b

7.8GB · Q4_K_M
Offline
JL3 Vigil·16GB
codereasoning
Avg Response
8.9s
VRAM
7.8GB

axe-sentinel-11b

6.5GB · Q4_K_M
Offline
JL3 Vigil·16GB
chatanalysis
Avg Response
7.2s
VRAM
6.5GB

axe-nova-22b

13.2GB · Q4_K_M
Offline
JL3 Vigil·16GB
reasoninglong-context
Avg Response
11.5s
VRAM
13.2GB

axe-muse-15b

9.1GB · Q4_K_M
Offline
JL1 Nova·64GB
creativechat
Avg Response
9.8s
VRAM
9.1GB

axe-prism-18b

10.8GB · Q4_K_M
Offline
JL3 Vigil·16GB
multimodalvision
Avg Response
10.2s
VRAM
10.8GB

Heavy (15s+)(6)

axe-titan-57b

34.1GB · Q4_K_M
Offline
JL1 Nova·64GB
complex-reasoningarchitecture
Avg Response
18.5s
VRAM
34.1GB

axe-nexus-70b

41.4GB · Q4_K_M
Offline
JL1 Nova·64GB
everything
Avg Response
25.0s
VRAM
41.4GB

axe-cortex-72b

43.2GB · Q4_K_M
Offline
JL1 Nova·64GB
reasoningresearchanalysis
Avg Response
26.3s
VRAM
43.2GB

axe-coder-34b

20.5GB · Q4_K_M
Offline
JL1 Nova·64GB
codearchitecture
Avg Response
15.2s
VRAM
20.5GB

axe-guardian-32b

19.3GB · Q4_K_M
Offline
JL1 Nova·64GB
analysisstrategy
Avg Response
14.7s
VRAM
19.3GB

axe-matrix-40b

24.1GB · Q4_K_M
Offline
JL1 Nova·64GB
reasoningresearch
Avg Response
17.1s
VRAM
24.1GB

Cloud(4)

Claude Sonnet 4.5

Anthropic
Online
Cloud·∞
codereasoningcreative
Avg Response
2.1s

Gemini Pro 1.5

Google
Online
Cloud·∞
multimodalchatlong-context
Avg Response
1.8s

Grok Beta

xAI
Online
Cloud·∞
analysisresearch
Avg Response
2.5s

Groq Llama 3.3

Groq
Online
Cloud·∞
fast-chatreasoning
Avg Response
0.8s