Model
Fit
Speed
Memory
Scores / Run
Runs comfortably
Ornith-1.5-35B-A3B
Q6_K
29.2GB
262k ctx
Runs comfortably
405
tok/s
ⓘ estimated
29
/ 32GB
91
/
89
/
86
Runs comfortably
405 tok/s
ⓘ estimated
29/32GB
Ornith-1.5-35B-A3B
Q5_K_M
25.3GB
262k ctx
Runs comfortably
467
tok/s
ⓘ estimated
25
/ 32GB
91
/
89
/
86
Runs comfortably
467 tok/s
ⓘ estimated
25/32GB
Ornith-1.5-35B-A3B
Q4_K_M
21.7GB
262k ctx
Runs comfortably
544
tok/s
ⓘ estimated
22
/ 32GB
91
/
89
/
86
Runs comfortably
544 tok/s
ⓘ estimated
22/32GB
Qwen3.6 35B A3B
Q4_K_M
20.0GB
262k ctx
Runs comfortably
574
tok/s
ⓘ estimated
20
/ 32GB
90
/
91
/
87
Runs comfortably
574 tok/s
ⓘ estimated
20/32GB
Pokee-Isaac 28B
Q4_K_M
16.0GB
10000k ctx
Runs comfortably
62
tok/s
ⓘ estimated
16
/ 32GB
90
/
88
/
88
Runs comfortably
62 tok/s
ⓘ estimated
16/32GB
Runs comfortably
35 tok/s
ⓘ estimated
28/32GB
17.6GB
262k ctx
Runs comfortably
56
tok/s
ⓘ estimated
18
/ 32GB
88
/
89
/
85
Runs comfortably
56 tok/s
ⓘ estimated
18/32GB
9.8GB
262k ctx
Runs comfortably
100
tok/s
ⓘ estimated
10
/ 32GB
88
/
89
/
85
Runs comfortably
100 tok/s
ⓘ estimated
10/32GB
6.2GB
262k ctx
Runs comfortably
159
tok/s
ⓘ estimated
6
/ 32GB
88
/
89
/
85
Runs comfortably
159 tok/s
ⓘ estimated
6/32GB
15.2GB
262k ctx
Runs comfortably
65
tok/s
ⓘ estimated
15
/ 32GB
88
/
89
/
85
Runs comfortably
65 tok/s
ⓘ estimated
15/32GB
Runs comfortably
54 tok/s
ⓘ estimated
18/32GB
Runs comfortably
55 tok/s
ⓘ estimated
18/32GB
Runs comfortably
62 tok/s
ⓘ estimated
16/32GB
Gemma 4 26B A4B
Q4_K_M
13.0GB
256k ctx
Runs comfortably
502
tok/s
ⓘ estimated
13
/ 32GB
85
/
88
/
84
Runs comfortably
502 tok/s
ⓘ estimated
13/32GB
Runs comfortably
466 tok/s
ⓘ estimated
14/32GB
Runs comfortably
261 tok/s
ⓘ estimated
25/32GB
Runs comfortably
35 tok/s
ⓘ estimated
28/32GB
Runs comfortably
60 tok/s
ⓘ estimated
17/32GB
Runs comfortably
35 tok/s
ⓘ estimated
29/32GB
Runs comfortably
40 tok/s
ⓘ estimated
25/32GB
Runs comfortably
49 tok/s
ⓘ estimated
20/32GB
Runs comfortably
148 tok/s
ⓘ estimated
7/32GB
Runs comfortably
130 tok/s
ⓘ estimated
8/32GB
Runs comfortably
101 tok/s
ⓘ estimated
10/32GB
Runs comfortably
54 tok/s
ⓘ estimated
18/32GB
Runs comfortably
170 tok/s
ⓘ estimated
6/32GB
Runs comfortably
646 tok/s
ⓘ estimated
16/32GB
Runs comfortably
34 tok/s
ⓘ estimated
29/32GB
Runs comfortably
66 tok/s
ⓘ estimated
15/32GB
Runs comfortably
116 tok/s
ⓘ estimated
9/32GB
Xing4.0-29B-A4B
Q4_K_M
20.1GB
256k ctx
Runs comfortably
355
tok/s
ⓘ estimated
20
/ 32GB
75
/
73
/
72
Runs comfortably
355 tok/s
ⓘ estimated
20/32GB
21.5GB
128k ctx
Runs comfortably
46
tok/s
ⓘ estimated
22
/ 32GB
72
/
77
/
80
Runs comfortably
46 tok/s
ⓘ estimated
22/32GB
Runs comfortably
66 tok/s
ⓘ estimated
15/32GB
Runs comfortably
109 tok/s
ⓘ estimated
9/32GB
Runs comfortably
76 tok/s
ⓘ estimated
13/32GB
Runs comfortably
137 tok/s
ⓘ estimated
7/32GB
Runs comfortably
58 tok/s
ⓘ estimated
17/32GB
Runs comfortably
35 tok/s
ⓘ estimated
28/32GB
Runs comfortably
116 tok/s
ⓘ estimated
9/32GB
Runs comfortably
189 tok/s
ⓘ estimated
5/32GB
Runs comfortably
73 tok/s
ⓘ estimated
14/32GB
Mixtral 8x7B Instruct
Q4_K_M
26.5GB
32k ctx
Runs comfortably
135
tok/s
ⓘ estimated
27
/ 32GB
62
/
65
/
68
Runs comfortably
135 tok/s
ⓘ estimated
27/32GB
Mixtral 8x7B Instruct
Q3_K_M
19.0GB
32k ctx
Runs comfortably
188
tok/s
ⓘ estimated
19
/ 32GB
62
/
65
/
68
Runs comfortably
188 tok/s
ⓘ estimated
19/32GB
Runs comfortably
131 tok/s
ⓘ estimated
8/32GB
Runs comfortably
75 tok/s
ⓘ estimated
13/32GB
Runs comfortably
135 tok/s
ⓘ estimated
7/32GB
Runs comfortably
379 tok/s
ⓘ estimated
3/32GB
Runs comfortably
394 tok/s
ⓘ estimated
3/32GB
Runs comfortably
240 tok/s
ⓘ estimated
4/32GB
16.0GB
128k ctx
Runs comfortably
62
tok/s
ⓘ estimated
16
/ 32GB
52
/
55
/
62
Runs comfortably
62 tok/s
ⓘ estimated
16/32GB
8.0GB
128k ctx
Runs comfortably
123
tok/s
ⓘ estimated
8
/ 32GB
52
/
55
/
62
Runs comfortably
123 tok/s
ⓘ estimated
8/32GB
Llama 3.1 8B Instruct
Q4_K_M
4.5GB
128k ctx
Runs comfortably
219
tok/s
ⓘ estimated
5
/ 32GB
52
/
55
/
62
Runs comfortably
219 tok/s
ⓘ estimated
5/32GB
Runs comfortably
197 tok/s
ⓘ estimated
5/32GB
Runs comfortably
328 tok/s
ⓘ estimated
3/32GB
3.5GB
128k ctx
Runs comfortably
282
tok/s
ⓘ estimated
4
/ 32GB
35
/
38
/
50
Runs comfortably
282 tok/s
ⓘ estimated
4/32GB
Llama 3.2 3B Instruct
Q4_K_M
2.0GB
128k ctx
Runs comfortably
493
tok/s
ⓘ estimated
2
/ 32GB
35
/
38
/
50
Runs comfortably
493 tok/s
ⓘ estimated
2/32GB
Runs tight
Runs tight
323 tok/s
ⓘ estimated
31/32GB
Llama 3.3 70B Instruct
Q3_K_M
30.0GB
128k ctx
Runs tight
33
tok/s
ⓘ estimated
30
/ 32GB
72
/
77
/
80
Runs tight
33 tok/s
ⓘ estimated
30/32GB
Runs tight
55 tok/s
ⓘ estimated
18/32GB
Runs with CPU offload
Qwen3.8-Flash-Next
EXL3_3BPW
85.0GB
262k ctx
Runs with CPU offload
348
tok/s
ⓘ estimated
85
/ 32GB
92
/
92
/
89
Runs with CPU offload
348 tok/s
ⓘ estimated
85/32GB
Qwen3.8-Flash-Next
EXL3_2BPW
62.7GB
262k ctx
Runs with CPU offload
471
tok/s
ⓘ estimated
63
/ 32GB
92
/
92
/
89
Runs with CPU offload
471 tok/s
ⓘ estimated
63/32GB
Ornith-1.5-35B-A3B
Q8_0
37.8GB
0k cap
Runs with CPU offload
313
tok/s
ⓘ estimated
38
/ 32GB
91
/
89
/
86
Runs with CPU offload
313 tok/s
ⓘ estimated
38/32GB
Qwen3.6 35B A3B
Q8_0
37.0GB
0k cap
Runs with CPU offload
311
tok/s
ⓘ estimated
37
/ 32GB
90
/
91
/
87
Runs with CPU offload
311 tok/s
ⓘ estimated
37/32GB
Runs with CPU offload
30 tok/s
ⓘ estimated
33/32GB
Gemma 4 26B A4B
BF16
50.0GB
0k cap
Runs with CPU offload
131
tok/s
ⓘ estimated
50
/ 32GB
85
/
88
/
84
Runs with CPU offload
131 tok/s
ⓘ estimated
50/32GB
Runs with CPU offload
29 tok/s
ⓘ estimated
34/32GB
Nemotron 3.5 Lightning
Q4_K_M
32.0GB
0k cap
Runs with CPU offload
270
tok/s
ⓘ estimated
32
/ 32GB
78
/
76
/
74
Runs with CPU offload
270 tok/s
ⓘ estimated
32/32GB
Llama 3.3 70B Instruct
Q4_K_M
42.0GB
0k cap
Runs with CPU offload
23
tok/s
ⓘ estimated
42
/ 32GB
72
/
77
/
80
Runs with CPU offload
23 tok/s
ⓘ estimated
42/32GB
Runs with CPU offload
30 tok/s
ⓘ estimated
33/32GB
Coding alternatives
Hosted coding assistants for comparison - per seat
Cursor
Hobby
AI-first code editor. Uses frontier models (GPT-4, Claude). $20/user Business...
$0.0
GitHub Copilot
Individual
AI coding assistant by GitHub. GPT-4o and Claude-based completions. $19/seat ...
$10.0
GitHub Copilot
Business
AI coding assistant by GitHub. GPT-4o and Claude-based completions. $19/seat ...
$19.0
Cursor
Pro
AI-first code editor. Uses frontier models (GPT-4, Claude). $20/user Business...
$20.0
GitHub Copilot
Enterprise
AI coding assistant by GitHub. GPT-4o and Claude-based completions. $19/seat ...
$39.0
Cursor
Business
AI-first code editor. Uses frontier models (GPT-4, Claude). $20/user Business...
$40.0
Per-token monthly figures are estimates at typical usage. Prices verified periodically - always confirm current pricing on the provider's website.
Sign in with GitHub to save this rig and get weekly model updates
Sign in with GitHub