Dataset Viewer
Auto-converted to Parquet Duplicate
date
stringdate
2026-08-19 00:00:00
2026-09-09 00:00:00
provider
stringclasses
2 values
gpu_model
stringclasses
114 values
gpu_model_key
stringclasses
112 values
vram_gb
int64
4
288
variant
stringclasses
4 values
price_usd_per_hour
float64
0
12.5
offer_count
float64
1
51
2026-08-19
runpod
A100 PCIe
A100 PCIE
80
community
1.19
null
2026-08-19
runpod
A100 PCIe
A100 PCIE
80
secure
1.39
null
2026-08-19
runpod
A100 SXM
A100 SXM
80
community
1.39
null
2026-08-19
runpod
A100 SXM
A100 SXM
80
secure
1.59
null
2026-08-19
runpod
A100 SXM 40GB
A100 SXM 40GB
40
community
1
null
2026-08-19
runpod
A40
A40
48
secure
0.44
null
2026-08-19
runpod
B200
B200
180
community
5.98
null
2026-08-19
runpod
B200
B200
180
secure
6.79
null
2026-08-19
runpod
B300
B300
288
community
6.94
null
2026-08-19
runpod
B300
B300
288
secure
7.89
null
2026-08-19
runpod
H100 NVL
H100 NVL
94
community
2.59
null
2026-08-19
runpod
H100 NVL
H100 NVL
94
secure
3.19
null
2026-08-19
runpod
H100 PCIe
H100 PCIE
80
community
1.99
null
2026-08-19
runpod
H100 PCIe
H100 PCIE
80
secure
2.89
null
2026-08-19
runpod
H100 SXM
H100 SXM
80
community
2.69
null
2026-08-19
runpod
H100 SXM
H100 SXM
80
secure
3.29
null
2026-08-19
runpod
H200 NVL
H200 NVL
143
secure
3.79
null
2026-08-19
runpod
H200 SXM
H200 SXM
141
community
3.59
null
2026-08-19
runpod
H200 SXM
H200 SXM
141
secure
4.59
null
2026-08-19
runpod
L4
L4
24
secure
0.49
null
2026-08-19
runpod
L40
L40
48
community
0.69
null
2026-08-19
runpod
L40
L40
48
secure
0.82
null
2026-08-19
runpod
L40S
L40S
48
community
0.79
null
2026-08-19
runpod
L40S
L40S
48
secure
0.99
null
2026-08-19
runpod
MI300X
MI300X
192
secure
2.39
null
2026-08-19
runpod
PRO 6000 MIG 24GB
PRO 6000 MIG 24GB
24
secure
0.59
null
2026-08-19
runpod
PRO 6000 MIG 48GB
PRO 6000 MIG 48GB
48
secure
1.09
null
2026-08-19
runpod
RTX 2000 Ada
RTX 2000 ADA
16
secure
0.24
null
2026-08-19
runpod
RTX 3070
RTX 3070
8
community
0.13
null
2026-08-19
runpod
RTX 3080
RTX 3080
10
community
0.17
null
2026-08-19
runpod
RTX 3080 Ti
RTX 3080 TI
12
community
0.18
null
2026-08-19
runpod
RTX 3090
RTX 3090
24
community
0.22
null
2026-08-19
runpod
RTX 3090
RTX 3090
24
secure
0.5
null
2026-08-19
runpod
RTX 3090 Ti
RTX 3090 TI
24
community
0.27
null
2026-08-19
runpod
RTX 4000 Ada
RTX 4000 ADA
20
community
0.2
null
2026-08-19
runpod
RTX 4000 Ada
RTX 4000 ADA
20
secure
0.28
null
2026-08-19
runpod
RTX 4000 Ada SFF
RTX 4000 ADA SFF
20
community
0.18
null
2026-08-19
runpod
RTX 4070 Ti
RTX 4070 TI
12
community
0.19
null
2026-08-19
runpod
RTX 4080
RTX 4080
16
community
0.27
null
2026-08-19
runpod
RTX 4080 SUPER
RTX 4080 SUPER
16
community
0.28
null
2026-08-19
runpod
RTX 4090
RTX 4090
24
community
0.34
null
2026-08-19
runpod
RTX 4090
RTX 4090
24
secure
0.74
null
2026-08-19
runpod
RTX 5000 Ada
RTX 5000 ADA
32
community
0.49
null
2026-08-19
runpod
RTX 5080
RTX 5080
16
community
0.39
null
2026-08-19
runpod
RTX 5090
RTX 5090
32
community
0.69
null
2026-08-19
runpod
RTX 5090
RTX 5090
32
secure
0.99
null
2026-08-19
runpod
RTX 6000 Ada
RTX 6000 ADA
48
community
0.74
null
2026-08-19
runpod
RTX 6000 Ada
RTX 6000 ADA
48
secure
0.84
null
2026-08-19
runpod
RTX A2000
RTX A2000
6
community
0.12
null
2026-08-19
runpod
RTX A4000
RTX A4000
16
community
0.17
null
2026-08-19
runpod
RTX A4000
RTX A4000
16
secure
0.25
null
2026-08-19
runpod
RTX A4500
RTX A4500
20
community
0.19
null
2026-08-19
runpod
RTX A4500
RTX A4500
20
secure
0.25
null
2026-08-19
runpod
RTX A5000
RTX A5000
24
community
0.16
null
2026-08-19
runpod
RTX A5000
RTX A5000
24
secure
0.27
null
2026-08-19
runpod
RTX A6000
RTX A6000
48
community
0.33
null
2026-08-19
runpod
RTX A6000
RTX A6000
48
secure
0.53
null
2026-08-19
runpod
RTX PRO 4000
RTX PRO 4000
24
secure
0.57
null
2026-08-19
runpod
RTX PRO 4500
RTX PRO 4500
32
secure
0.72
null
2026-08-19
runpod
RTX PRO 4500 SE
RTX PRO 4500 SE
32
secure
0.72
null
2026-08-19
runpod
RTX PRO 5000
RTX PRO 5000
48
community
0.82
null
2026-08-19
runpod
RTX PRO 5000
RTX PRO 5000
48
secure
0.96
null
2026-08-19
runpod
RTX PRO 6000
RTX PRO 6000
96
community
1.69
null
2026-08-19
runpod
RTX PRO 6000
RTX PRO 6000
96
secure
2.09
null
2026-08-19
runpod
RTX PRO 6000 MaxQ
RTX PRO 6000 MAXQ
96
community
1.64
null
2026-08-19
runpod
RTX PRO 6000 WK
RTX PRO 6000 WK
96
community
1.69
null
2026-08-19
runpod
RTX PRO 6000 WK
RTX PRO 6000 WK
96
secure
1.89
null
2026-08-19
runpod
Tesla V100
TESLA V100
16
community
0.19
null
2026-08-19
runpod
V100 SXM2
V100 SXM2
16
community
0.23
null
2026-08-19
vast.ai
A10
A10
22
median
0.2689
1
2026-08-19
vast.ai
A10
A10
22
min
0.2689
1
2026-08-19
vast.ai
A100 PCIE
A100 PCIE
40
median
0.8674
15
2026-08-19
vast.ai
A100 PCIE
A100 PCIE
40
min
0.6563
15
2026-08-19
vast.ai
A100 SXM4
A100 SXM4
80
median
0.8348
11
2026-08-19
vast.ai
A100 SXM4
A100 SXM4
80
min
0.2689
11
2026-08-19
vast.ai
A800 PCIE
A800 PCIE
80
median
0.5676
6
2026-08-19
vast.ai
A800 PCIE
A800 PCIE
80
min
0.4674
6
2026-08-19
vast.ai
B200
B200
179
median
6.7943
9
2026-08-19
vast.ai
B200
B200
179
min
5.3146
9
2026-08-19
vast.ai
B300
B300
269
median
11.2507
1
2026-08-19
vast.ai
B300
B300
269
min
11.2507
1
2026-08-19
vast.ai
CMP 170HX
CMP 170HX
64
median
0.2684
6
2026-08-19
vast.ai
CMP 170HX
CMP 170HX
64
min
0.1889
6
2026-08-19
vast.ai
GTX 1050 Ti
GTX 1050 TI
4
median
0.6111
1
2026-08-19
vast.ai
GTX 1050 Ti
GTX 1050 TI
4
min
0.6111
1
2026-08-19
vast.ai
GTX 1060
GTX 1060
6
median
0.0881
5
2026-08-19
vast.ai
GTX 1060
GTX 1060
6
min
0.0193
5
2026-08-19
vast.ai
GTX 1070
GTX 1070
8
median
0.0484
8
2026-08-19
vast.ai
GTX 1070
GTX 1070
8
min
0.0478
8
2026-08-19
vast.ai
GTX 1070 Ti
GTX 1070 TI
8
median
0.0529
6
2026-08-19
vast.ai
GTX 1070 Ti
GTX 1070 TI
8
min
0.051
6
2026-08-19
vast.ai
GTX 1080
GTX 1080
8
median
0.0529
23
2026-08-19
vast.ai
GTX 1080
GTX 1080
8
min
0.0278
23
2026-08-19
vast.ai
GTX 1080 Ti
GTX 1080 TI
11
median
0.0657
17
2026-08-19
vast.ai
GTX 1080 Ti
GTX 1080 TI
11
min
0.0556
17
2026-08-19
vast.ai
GTX 1650
GTX 1650
4
median
0.2778
1
2026-08-19
vast.ai
GTX 1650
GTX 1650
4
min
0.2778
1
2026-08-19
vast.ai
GTX 1650 S
GTX 1650 S
4
median
0.6778
1
2026-08-19
vast.ai
GTX 1650 S
GTX 1650 S
4
min
0.6778
1
2026-08-19
vast.ai
GTX 1660
GTX 1660
6
median
0.6778
1
End of preview. Expand in Data Studio

AI compute prices, collected daily

Two tables, updated every day by overfeed.news:

  • gpu_prices.csv — hourly rental price of every GPU listed on RunPod (catalogue: secure, community) and Vast.ai (marketplace: min, median, with offer_count), one row per day, provider, chip and tier.
  • model_prices.csv — price per million tokens (input and output) of every model served through OpenRouter, one row per day and model, with context length, parameter count and Hugging Face id when the weights are open.

Last update: 2026-09-09. GPU table: 4930 rows over 22 days. Model table: 8045 rows over 23 days.

How it is measured

Prices are read once a day from the providers' public APIs, stored as published, and never smoothed. A Vast.ai min is the cheapest listing at collection time and can be a single machine; median with offer_count is the number to compare against. Details, and the same data as pages with 90-day charts: https://overfeed.news/gpus and https://overfeed.news/model-prices. Method: https://overfeed.news/guides/how-we-measure.

License

CC BY 4.0. Cite as "overfeed.news AI compute prices" with a link to https://overfeed.news.

Downloads last month
-