Hugging Face's logo

Datasets:
sambombay
/
ai-model-price-changes

Languages:
English
Size:
n<1K
Tags:
ai
pricing
llm
economics
License:
Dataset card Data Studio Files
xet
 Community
Dataset Preview
Duplicate
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed because of a cast error
Error code:   DatasetGenerationCastError
Exception:    DatasetGenerationCastError
Message:      An error occurred while generating the dataset

All the data files must have the same columns, but at some point there are 3 new columns ({'input_per_1m_usd', 'context_window', 'output_per_1m_usd'}) and 5 missing columns ({'pct_change', 'date', 'price_to', 'price_from', 'field'}).

This happened while the csv dataset builder was generating data using

hf://datasets/sambombay/ai-model-price-changes/prices.csv (at revision 18afd81b5e09fa070578a56421c8f9abf468f869), ['hf://datasets/sambombay/ai-model-price-changes@18afd81b5e09fa070578a56421c8f9abf468f869/changes.csv', 'hf://datasets/sambombay/ai-model-price-changes@18afd81b5e09fa070578a56421c8f9abf468f869/prices.csv']

Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1856, in _prepare_split_single
                  writer.write_table(table)
                  ~~~~~~~~~~~~~~~~~~^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 811, in write_table
                  self._write_table(pa_table, writer_batch_size=writer_batch_size)
                  ~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 822, in _write_table
                  pa_table = table_cast(pa_table, self._schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2381, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2309, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              model_id: string
              model_name: string
              input_per_1m_usd: double
              output_per_1m_usd: double
              context_window: int64
              -- schema metadata --
              pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 908
              to
              {'date': Value('string'), 'model_id': Value('string'), 'model_name': Value('string'), 'field': Value('string'), 'price_from': Value('float64'), 'price_to': Value('float64'), 'pct_change': Value('float64')}
              because column names don't match
              
              During handling of the above exception, another exception occurred:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1379, in compute_config_parquet_and_info_response
                  parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
                                                                        ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      builder, max_dataset_size_bytes=max_dataset_size_bytes
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 956, in stream_convert_to_parquet
                  builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
                  ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1702, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                                               ~~~~~~~~~~~~~~~~~~~~~~~~~~^
                      gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  ):
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1858, in _prepare_split_single
                  raise DatasetGenerationCastError.from_cast_error(
                  ...<4 lines>...
                  )
              datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
              
              All the data files must have the same columns, but at some point there are 3 new columns ({'input_per_1m_usd', 'context_window', 'output_per_1m_usd'}) and 5 missing columns ({'pct_change', 'date', 'price_to', 'price_from', 'field'}).
              
              This happened while the csv dataset builder was generating data using
              
              hf://datasets/sambombay/ai-model-price-changes/prices.csv (at revision 18afd81b5e09fa070578a56421c8f9abf468f869), ['hf://datasets/sambombay/ai-model-price-changes@18afd81b5e09fa070578a56421c8f9abf468f869/changes.csv', 'hf://datasets/sambombay/ai-model-price-changes@18afd81b5e09fa070578a56421c8f9abf468f869/prices.csv']
              
              Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

date
string
model_id
string
model_name
string
field
string
price_from
float64
price_to
float64
pct_change
float64
2026-10-08
~deepseek/deepseek-flash-latest
DeepSeek: DeepSeek Flash Latest
input
0.019
0.0283
48.9
2026-10-08
~deepseek/deepseek-flash-latest
DeepSeek: DeepSeek Flash Latest
output
0.424
1
135.8
2026-10-08
~deepseek/deepseek-pro-latest
DeepSeek: DeepSeek Pro Latest
input
0.2856
0.108
-62.2
2026-10-08
~deepseek/deepseek-pro-latest
DeepSeek: DeepSeek Pro Latest
output
4.2
8
90.5
2026-10-08
deepseek/deepseek-v4-flash
DeepSeek: DeepSeek V4 Flash 0423
input
0.03
0.0099
-67
2026-10-08
deepseek/deepseek-v4-flash
DeepSeek: DeepSeek V4 Flash 0423
input
0.0099
0.0075
-24.2
2026-10-08
deepseek/deepseek-v4-flash-0731
DeepSeek: DeepSeek V4 Flash 0731
input
0.018
0.006
-66.7
2026-10-08
deepseek/deepseek-v4-flash-0731
DeepSeek: DeepSeek V4 Flash 0731
input
0.006
0.0046
-23.3
2026-10-08
~deepseek/deepseek-v4-flash-latest
DeepSeek: DeepSeek V4 Flash Latest
input
0.018
0.006
-66.7
2026-10-08
~deepseek/deepseek-v4-flash-latest
DeepSeek: DeepSeek V4 Flash Latest
input
0.006
0.0046
-23.3
2026-10-08
deepseek/deepseek-v4-pro
DeepSeek: DeepSeek V4 Pro 0423
input
0.2088
0.2923
40
2026-10-08
deepseek/deepseek-v4-pro
DeepSeek: DeepSeek V4 Pro 0423
output
0.4176
0.5846
40
2026-10-08
google/gemma-4-26b-a4b-it
Google: Gemma 4 26B A4B
input
0.0765
0.09
17.6
2026-10-08
google/gemma-4-26b-a4b-it
Google: Gemma 4 26B A4B
output
0.255
0.3
17.6
2026-10-08
meta-llama/llama-3.3-70b-instruct
Meta: Llama 3.3 70B Instruct
input
0.22
0.1
-54.5
2026-10-08
meta-llama/llama-3.3-70b-instruct
Meta: Llama 3.3 70B Instruct
output
0.5
0.32
-36
2026-10-08
moonshotai/kimi-k2.6
MoonshotAI: Kimi K2.6
input
0.465
0.4375
-5.9
2026-10-08
moonshotai/kimi-k3
MoonshotAI: Kimi K3
input
0.5
0.62
24
2026-10-08
moonshotai/kimi-k3
MoonshotAI: Kimi K3
output
15
12.3
-18
2026-10-08
~moonshotai/kimi-latest
MoonshotAI: Kimi Latest
input
0.49
0.45
-8.2
2026-10-08
nvidia/nemotron-3-nano-30b-a3b
NVIDIA: Nemotron 3 Nano 30B A3B
input
0.05
0.06
20
2026-10-08
nvidia/nemotron-3-nano-30b-a3b
NVIDIA: Nemotron 3 Nano 30B A3B
output
0.2
0.24
20
2026-10-08
nvidia/nemotron-3-super-120b-a12b
NVIDIA: Nemotron 3 Super
input
0.08
0.085
6.3
2026-10-08
nvidia/nemotron-3-super-120b-a12b
NVIDIA: Nemotron 3 Super
output
0.45
0.4
-11.1
2026-10-08
qwen/qwen3-30b-a3b-instruct-2507
Qwen: Qwen3 30B A3B Instruct 2507
input
0.0481
0.1
107.9
2026-10-08
qwen/qwen3-30b-a3b-instruct-2507
Qwen: Qwen3 30B A3B Instruct 2507
output
0.193
0.3
55.4
2026-10-08
qwen/qwen3-next-80b-a3b-instruct
Qwen: Qwen3 Next 80B A3B Instruct
input
0.1
0.15
50
2026-10-08
qwen/qwen3-next-80b-a3b-instruct
Qwen: Qwen3 Next 80B A3B Instruct
output
1.1
1.5
36.4
2026-10-08
qwen/qwen3.5-27b
Qwen: Qwen3.5-27B
input
0.195
0.26
33.3
2026-10-08
qwen/qwen3.5-27b
Qwen: Qwen3.5-27B
output
1.56
2.6
66.7
2026-10-08
qwen/qwen3.6-27b
Qwen: Qwen3.6 27B
input
0.3
0.45
50
2026-10-08
qwen/qwen3.6-27b
Qwen: Qwen3.6 27B
output
2
2.7
35
2026-10-08
stepfun/step-5-preview
StepFun: Step 5 Preview
added
null
2.7
null
2026-10-08
tencent/hy3
Tencent: Hy3
input
0.0825
0.132
60
2026-10-08
tencent/hy3
Tencent: Hy3
output
0.33
0.528
60
2026-10-08
tencent/hy4-preview
Tencent: Hy4 preview
input
0.7506
0.834
11.1
2026-10-08
tencent/hy4-preview
Tencent: Hy4 preview
output
2.2509
2.501
11.1
2026-10-08
z-ai/glm-5.2
Z.ai: GLM 5.2
input
0.171
0.03
-82.5
2026-10-08
z-ai/glm-5.2
Z.ai: GLM 5.2
output
7.2
10
38.9
2026-10-08
z-ai/glm-5.3
Z.ai: GLM 5.3
input
0.062
0.059
-4.8
2026-10-08
~z-ai/glm-flash-latest
Z.ai: GLM Flash Latest
input
0.04
0.032
-20
2026-10-08
~z-ai/glm-flash-latest
Z.ai: GLM Flash Latest
output
0.5
4.4401
788
2026-10-08
~z-ai/glm-latest
Z.ai: GLM Latest
input
0.062
0.0472
-23.9
2026-10-08
~z-ai/glm-latest
Z.ai: GLM Latest
output
3.39
12
254
2026-10-09
baidu/ernie-4.5-vl-424b-a47b
Baidu: ERNIE 4.5 VL 424B A47B
retired
1.25
null
null
2026-10-09
~deepseek/deepseek-flash-latest
DeepSeek: DeepSeek Flash Latest
input
0.0283
0.016
-43.5
2026-10-09
~deepseek/deepseek-flash-latest
DeepSeek: DeepSeek Flash Latest
output
1
1.2
20
2026-10-09
~deepseek/deepseek-pro-latest
DeepSeek: DeepSeek Pro Latest
input
0.108
0.1282
18.7
2026-10-09
deepseek/deepseek-chat
DeepSeek: DeepSeek V3
input
0.2574
0.32
24.3
2026-10-09
deepseek/deepseek-chat
DeepSeek: DeepSeek V3
output
1.0287
0.89
-13.5
2026-10-09
deepseek/deepseek-v4-flash
DeepSeek: DeepSeek V4 Flash 0423
input
0.0075
0.0099
32
2026-10-09
deepseek/deepseek-v4-flash
DeepSeek: DeepSeek V4 Flash 0423
input
0.0099
0.0075
-24.2
2026-10-09
deepseek/deepseek-v4-flash-0731
DeepSeek: DeepSeek V4 Flash 0731
input
0.0046
0.018
291.3
2026-10-09
deepseek/deepseek-v4-flash-0731
DeepSeek: DeepSeek V4 Flash 0731
input
0.018
0.0137
-23.9
2026-10-09
~deepseek/deepseek-v4-flash-latest
DeepSeek: DeepSeek V4 Flash Latest
input
0.0046
0.0144
213
2026-10-09
~deepseek/deepseek-v4-flash-latest
DeepSeek: DeepSeek V4 Flash Latest
output
1.28
0.0341
-97.3
2026-10-09
~deepseek/deepseek-v4-flash-latest
DeepSeek: DeepSeek V4 Flash Latest
input
0.0144
0.011
-23.6
2026-10-09
~deepseek/deepseek-v4-flash-latest
DeepSeek: DeepSeek V4 Flash Latest
output
0.0341
0.0307
-10
2026-10-09
deepseek/deepseek-v4-pro
DeepSeek: DeepSeek V4 Pro 0423
input
0.2923
0.2871
-1.8
2026-10-09
deepseek/deepseek-v4-pro
DeepSeek: DeepSeek V4 Pro 0423
output
0.5846
0.5742
-1.8
2026-10-09
meta-llama/llama-3.3-70b-instruct
Meta: Llama 3.3 70B Instruct
input
0.1
0.22
120
2026-10-09
meta-llama/llama-3.3-70b-instruct
Meta: Llama 3.3 70B Instruct
output
0.32
0.5
56.2
2026-10-09
mistralai/mistral-nemo
Mistral: Mistral Nemo
input
0.019
0.029
52.6
2026-10-09
moonshotai/kimi-k2.6
MoonshotAI: Kimi K2.6
input
0.4375
0.4357
-0.4
2026-10-09
moonshotai/kimi-k3
MoonshotAI: Kimi K3
input
0.62
0.5
-19.4
2026-10-09
moonshotai/kimi-k3
MoonshotAI: Kimi K3
output
12.3
12
-2.4
2026-10-09
~moonshotai/kimi-latest
MoonshotAI: Kimi Latest
input
0.45
0.49
8.9
2026-10-09
nvidia/nemotron-3-super-120b-a12b
NVIDIA: Nemotron 3 Super
input
0.085
0.08
-5.9
2026-10-09
nvidia/nemotron-3-super-120b-a12b
NVIDIA: Nemotron 3 Super
output
0.4
0.45
12.5
2026-10-09
nvidia/nemotron-3.5-lightning
NVIDIA: Nemotron 3.5 Lightning
input
0.049
0.0469
-4.3
2026-10-09
nvidia/nemotron-3.5-lightning
NVIDIA: Nemotron 3.5 Lightning
output
0.14
0.134
-4.3
2026-10-09
openai/gpt-6.1-sol:batch
OpenAI: GPT-6.1 Sol (batch)
added
null
5
null
2026-10-09
openai/gpt-6.1-sol-pro:batch
OpenAI: GPT-6.1 Sol Pro (batch)
added
null
5
null
2026-10-09
qwen/qwen3-next-80b-a3b-instruct
Qwen: Qwen3 Next 80B A3B Instruct
input
0.15
0.1
-33.3
2026-10-09
qwen/qwen3-next-80b-a3b-instruct
Qwen: Qwen3 Next 80B A3B Instruct
output
1.5
1.1
-26.7
2026-10-09
qwen/qwen3.5-397b-a17b
Qwen: Qwen3.5 397B A17B
input
0.45
0.55
22.2
2026-10-09
qwen/qwen3.5-397b-a17b
Qwen: Qwen3.5 397B A17B
output
3
3.5
16.7
2026-10-09
qwen/qwen3.5-35b-a3b
Qwen: Qwen3.5-35B-A3B
input
0.15
0.08
-46.7
2026-10-09
qwen/qwen3.5-35b-a3b
Qwen: Qwen3.5-35B-A3B
output
1
0.75
-25
2026-10-09
z-ai/glm-5.2
Z.ai: GLM 5.2
input
0.03
0.06
100
2026-10-09
z-ai/glm-5.2
Z.ai: GLM 5.2
output
10
6
-40
2026-10-09
z-ai/glm-5.3
Z.ai: GLM 5.3
input
0.059
0.04
-32.2
2026-10-09
z-ai/glm-5.3
Z.ai: GLM 5.3
output
3.39
3.6
6.2
2026-10-09
~z-ai/glm-flash-latest
Z.ai: GLM Flash Latest
output
4.4401
0.3268
-92.6
2026-10-09
~z-ai/glm-latest
Z.ai: GLM Latest
input
0.0472
0.0312
-33.9
2026-10-09
~z-ai/glm-latest
Z.ai: GLM Latest
output
12
4.4
-63.3
2026-10-09
inclusionai/ling-3.0-flash-sante
inclusionAI: Ling 3.0 Flash Sante
added
null
0.1232
null
2026-10-09
inclusionai/ling-3.0-flash-sante:free
inclusionAI: Ling 3.0 Flash Sante (free)
retired
null
null
null
2026-10-10
~deepseek/deepseek-flash-latest
DeepSeek: DeepSeek Flash Latest
input
0.016
0.0001
-99.4
2026-10-10
~deepseek/deepseek-flash-latest
DeepSeek: DeepSeek Flash Latest
output
1.2
0.6
-50
2026-10-10
~deepseek/deepseek-pro-latest
DeepSeek: DeepSeek Pro Latest
input
0.1282
0.2191
70.9
2026-10-10
~deepseek/deepseek-pro-latest
DeepSeek: DeepSeek Pro Latest
output
8
0.6574
-91.8
2026-10-10
deepseek/deepseek-v3.2
DeepSeek: DeepSeek V3.2
output
0.42
0.8
90.5
2026-10-10
deepseek/deepseek-v4-flash
DeepSeek: DeepSeek V4 Flash 0423
input
0.0075
0.029
286.7
2026-10-10
deepseek/deepseek-v4-flash-0731
DeepSeek: DeepSeek V4 Flash 0731
input
0.0137
0.0154
12.4
2026-10-10
~deepseek/deepseek-v4-flash-latest
DeepSeek: DeepSeek V4 Flash Latest
input
0.011
0.0154
40
2026-10-10
~deepseek/deepseek-v4-flash-latest
DeepSeek: DeepSeek V4 Flash Latest
output
0.0307
1.28
4,069.4
2026-10-10
deepseek/deepseek-v4-pro
DeepSeek: DeepSeek V4 Pro 0423
input
0.2871
0.2347
-18.3
2026-10-10
deepseek/deepseek-v4-pro
DeepSeek: DeepSeek V4 Pro 0423
output
0.5742
0.4695
-18.2
2026-10-10
deepseek/deepseek-v4-pro
DeepSeek: DeepSeek V4 Pro 0423
input
0.2347
0.2279
-2.9
End of preview.

AI Model Price Changes

A daily log of retail price changes across the AI model market, derived from the public OpenRouter model catalogue. One row per price event: which model, which field, the price before, the price after.

Maintained by Sam Bombay / Noir & Co. Live view: https://ai-price-watch-beta.vercel.app

What is in it

  • changes.csv - every logged change: date, model_id, model_name, field, price_from, price_to, pct_change. field is input, output, added (new model) or retired (withdrawn). All prices are USD per 1,000,000 tokens.
  • prices.csv - the current snapshot: input and output price and context window per model.

Coverage: 800 events across 452 models, 2026-10-08 to 2026-11-01. Of 715 comparable price events, 365 are cuts and 349 are rises.

Why it is useful

AI list prices are not stable, and they do not all move in one direction. In this log the same model routinely falls more than 20 percent and climbs straight back, while other models are retired without notice. Anyone modelling AI unit economics, writing about AI costs, or auditing a vendor invoice needs the dated record rather than today's rate card.

Method and limits

  • Source: the public OpenRouter model catalogue, fetched once a day.
  • A change is logged when a model ID's input or output price differs from the previous fetch. New and retired model IDs are logged as added / retired.
  • Prices are retail list prices in USD per 1M tokens. They are not invoice prices; discounts, committed-volume rates and enterprise agreements are not visible here.
  • Catalogue aliases (for example a Latest pointer) are separate rows from the models they point at, and can move independently.
  • Dated from the first fetch, not from a vendor announcement. A change logged on date D means the difference was observed on D.

Citation

Bombay, S. (2026). AI Model Price Changes [Data set]. https://ai-price-watch-beta.vercel.app
Downloads last month
82
Total file size:
102 kB