Hugging Face's logo

Datasets:
Karmane
/
openrouter-model-pricing-capability-matrix

Tasks:
Tabular Classification
Modalities:
Tabular
Text
Formats:
parquet
Languages:
English
Size:
< 1K
Tags:
openrouter
llm-pricing
model-catalog
inference
api-pricing
procurement
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community

You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

This dataset requires a paid license. Please purchase access at the link below and provide your Hugging Face username. Once purchased, submit an access request on this page. Access is granted manually.

Log in or Sign Up to review the conditions and access this dataset content.

OpenRouter Model Pricing and Capability Matrix

This dataset packages 429 public OpenRouter model records across 75 developers into one analysis-ready table for AI procurement, model-selection workflows, router strategy, pricing research, and internal model-catalog dashboards.

It starts from OpenRouter's public models API and count endpoint, then enriches each model with its per-model endpoint metadata so buyers can compare token pricing, context windows, provider diversity, endpoint breadth, uptime coverage, reasoning flags, tool support, caching support, and modality mix without manually stitching the raw payloads together.

Current collection-time snapshot highlights:

  • 14.92% of listed models are free at the published OpenRouter price point
  • Multimodal, embedding, image, speech, video, and rerank outputs are all represented
  • Provider routing breadth is preserved so buyers can distinguish single-provider models from broader multi-provider options

Use cases

  • Shortlist cost-efficient OpenRouter models by blended token cost, context tier, and provider breadth
  • Filter for models that support tools, reasoning, structured outputs, or implicit caching
  • Build procurement dashboards for approved models, free-model scanning, or premium-route review
  • Compare open-weight Hugging Face-linked options against API-first closed models in one table

Dataset Preview

record_id model_id canonical_slug model_name developer_name developer_slug model_series_slug hugging_face_model_id hugging_face_model_url open_weights_available model_added_at model_added_date model_added_year model_age_days context_length top_provider_context_length top_provider_max_completion_tokens top_provider_is_moderated architecture_modality input_modalities_pipe output_modalities_pipe input_modality_count output_modality_count tokenizer instruct_type primary_market_segment context_tier supports_multimodal_input accepts_text_input accepts_image_input accepts_audio_input accepts_video_input accepts_file_input outputs_text outputs_image outputs_audio outputs_video outputs_embeddings outputs_transcription outputs_rerank knowledge_cutoff knowledge_cutoff_year expiration_date is_deprecated supported_parameters_pipe supported_parameter_count supports_tools supports_tool_choice supports_reasoning supports_include_reasoning supports_structured_outputs supports_response_format supports_temperature supports_top_p supports_top_k supports_seed supports_stop supports_frequency_penalty supports_presence_penalty supports_repetition_penalty default_parameters_json prompt_price_per_1m_tokens_usd completion_price_per_1m_tokens_usd request_price_usd image_input_price_usd audio_input_price_usd web_search_price_usd internal_reasoning_price_per_1m_tokens_usd input_cache_read_price_per_1m_tokens_usd input_cache_write_price_per_1m_tokens_usd blended_text_cost_per_1m_in_1m_out_usd price_tier is_free_model request_to_prompt_cost_ratio endpoint_count active_endpoint_count provider_count provider_diversity_tier provider_names_pipe provider_tags_pipe quantizations_pipe quantization_count has_region_specific_endpoints endpoint_context_length_min endpoint_context_length_max endpoint_max_completion_tokens_max endpoint_max_prompt_tokens_max best_uptime_last_1d_pct median_uptime_last_1d_pct supports_tools_endpoint_count supports_reasoning_endpoint_count supports_structured_outputs_endpoint_count supports_implicit_caching_endpoint_count per_request_limits_json supported_voices_pipe source_models_api_url source_model_details_api_url source_docs_url source_pricing_url source_domain last_collected_at
openrouter-nvidia_llama-nemotron-embed-vl-1b-v2_free nvidia/llama-nemotron-embed-vl-1b-v2:free nvidia/llama-nemotron-embed-vl-1b-v2-20260224 NVIDIA: Llama Nemotron Embed VL 1B V2 (free) NVIDIA nvidia llama-nemotron-embed-vl-1b-v2:free nvidia/llama-nemotron-embed-vl-1b-v2 https://huggingface.co/nvidia/llama-nemotron-embed-vl-1b-v2 True 2026-02-25T18:43:37+00:00 2026-02-25 2026 97 131072 131072 False text+image->embeddings text | image embeddings 2 1 Other embeddings extended-context True True True False False False False False False False True False False False max_tokens | seed | temperature | top_p 4 False False False False False False True True False True False False False False {"frequency_penalty": null, "temperature": null, "top_p": null} 0 0 False 0 free True False 0 0 0 single-provider 0 False 0 0 0 0 https://openrouter.ai/api/v1/models?output_modalities=all https://openrouter.ai/api/v1/models/nvidia/llama-nemotron-embed-vl-1b-v2-20260224/endpoints https://openrouter.ai/docs/guides/overview/models https://openrouter.ai/pricing openrouter.ai 2026-06-03T02:35:09.620431+00:00
openrouter-perplexity_pplx-embed-v1-0.6b perplexity/pplx-embed-v1-0.6b perplexity/pplx-embed-v1-0.6B Perplexity: Embed V1 0.6B Perplexity perplexity pplx-embed-v1-0.6b False 2026-03-16T01:34:28+00:00 2026-03-16 2026 79 32000 32000 False text->embeddings text embeddings 1 1 Other embeddings standard-context False True False False False False False False False False True False False False frequency_penalty | max_tokens | presence_penalty | temperature | top_k | top_p | web_search_o... [6 more chars] 7 False False False False False False True True True False False True True False {"frequency_penalty": null, "presence_penalty": null, "repetition_penalty": null, "temperature": nul... [32 more chars] 0.004 0 False 0.004 ultra-low False False 1 1 1 single-provider Perplexity perplexity/int8 int8 1 True 32000.0 32000.0 100.0 100.0 0 0 0 0 https://openrouter.ai/api/v1/models?output_modalities=all https://openrouter.ai/api/v1/models/perplexity/pplx-embed-v1-0.6B/endpoints https://openrouter.ai/docs/guides/overview/models https://openrouter.ai/pricing openrouter.ai 2026-06-03T02:35:09.620431+00:00
openrouter-baai_bge-base-en-v1.5 baai/bge-base-en-v1.5 baai/bge-base-en-v1.5-20251117 BAAI: bge-base-en-v1.5 BAAI baai bge-base-en-v1.5 BAAI/bge-base-en-v1.5 https://huggingface.co/BAAI/bge-base-en-v1.5 True 2025-11-18T02:10:37+00:00 2025-11-18 2025 197 8192 512 False text->embeddings text embeddings 1 1 Other embeddings short-context False True False False False False False False False False True False False False frequency_penalty | max_tokens | min_p | presence_penalty | repetition_penalty | response_forma... [50 more chars] 11 False False False False False True True True True True True True True True {"frequency_penalty": null, "temperature": null, "top_p": null} 0.005 0 False 0.005 ultra-low False False 1 1 1 single-provider DeepInfra deepinfra unknown 1 False 512.0 512.0 100.0 100.0 0 0 0 0 https://openrouter.ai/api/v1/models?output_modalities=all https://openrouter.ai/api/v1/models/baai/bge-base-en-v1.5-20251117/endpoints https://openrouter.ai/docs/guides/overview/models https://openrouter.ai/pricing openrouter.ai 2026-06-03T02:35:09.620431+00:00
openrouter-intfloat_e5-base-v2 intfloat/e5-base-v2 intfloat/e5-base-v2-20251117 Intfloat: E5-Base-v2 Intfloat intfloat e5-base-v2 intfloat/e5-base-v2 https://huggingface.co/intfloat/e5-base-v2 True 2025-11-18T02:33:12+00:00 2025-11-18 2025 197 8192 512 False text->embeddings text embeddings 1 1 Other embeddings short-context False True False False False False False False False False True False False False frequency_penalty | max_tokens | min_p | presence_penalty | repetition_penalty | response_forma... [50 more chars] 11 False False False False False True True True True True True True True True {"frequency_penalty": null, "temperature": null, "top_p": null} 0.005 0 False 0.005 ultra-low False False 1 1 1 single-provider DeepInfra deepinfra unknown 1 False 512.0 512.0 100.0 100.0 0 0 0 0 https://openrouter.ai/api/v1/models?output_modalities=all https://openrouter.ai/api/v1/models/intfloat/e5-base-v2-20251117/endpoints https://openrouter.ai/docs/guides/overview/models https://openrouter.ai/pricing openrouter.ai 2026-06-03T02:35:09.620431+00:00
openrouter-sentence-transformers_all-minilm-l12-v2 sentence-transformers/all-minilm-l12-v2 sentence-transformers/all-minilm-l12-v2-20251117 Sentence Transformers: all-MiniLM-L12-v2 Sentence Transformers sentence-transformers all-minilm-l12-v2 sentence-transformers/all-MiniLM-L12-v2 https://huggingface.co/sentence-transformers/all-MiniLM-L12-v2 True 2025-11-18T02:15:55+00:00 2025-11-18 2025 197 8192 512 False text->embeddings text embeddings 1 1 Other embeddings short-context False True False False False False False False False False True False False False frequency_penalty | max_tokens | min_p | presence_penalty | repetition_penalty | response_forma... [50 more chars] 11 False False False False False True True True True True True True True True {"frequency_penalty": null, "temperature": null, "top_p": null} 0.005 0 False 0.005 ultra-low False False 1 1 1 single-provider DeepInfra deepinfra unknown 1 False 512.0 512.0 100.0 100.0 0 0 0 0 https://openrouter.ai/api/v1/models?output_modalities=all https://openrouter.ai/api/v1/models/sentence-transformers/all-minilm-l12-v2-20251117/endpoints https://openrouter.ai/docs/guides/overview/models https://openrouter.ai/pricing openrouter.ai 2026-06-03T02:35:09.620431+00:00

Access Requirements (Paid Dataset)

This dataset is behind manual gated access.

To obtain access:

  1. Purchase the dataset here:
    https://thearticulated.gumroad.com/l/openrouter-model-pricing-capability-matrix

  2. Provide your Hugging Face username at checkout.

  3. Return to this Hugging Face page and click:
    "Request Access"

  4. Your access will be approved within 1-12 hours.

Once approved, you can use the Python snippet at the bottom of this README to load the dataset.

Dataset Structure

Total rows: 429

Total columns: 101

Splits

  • data: 429 rows

Data Files

  • data: data/data.parquet

Data Dictionary

The table below describes the columns included in this dataset.

column pandas_dtype dataset_type description
record_id object string Text column. Stable text identifier created from the OpenRouter model id.
model_id object string Text column. Current OpenRouter model identifier used in API requests.
canonical_slug object string Text column. Permanent OpenRouter canonical slug for the model.
model_name object string Text column. Human-readable model name returned by OpenRouter.
developer_name object string Text column. Normalized developer or vendor label derived from the model id and display name.
developer_slug object string Text column. Provider slug segment before the first slash in the OpenRouter model id.
model_series_slug object string Text column. Series slug segment after the first slash in the OpenRouter model id.
hugging_face_model_id object string Text column. Linked Hugging Face model id when OpenRouter exposes one.
hugging_face_model_url object string Text column. Direct Hugging Face model URL derived from the linked model id when available.
open_weights_available bool boolean True/false column. Boolean flag indicating that OpenRouter links the model to a Hugging Face model page.
model_added_at object datetime-like string Stored as text, but values appear to represent dates or timestamps. UTC timestamp when the model was added to OpenRouter.
model_added_date object datetime-like string Stored as text, but values appear to represent dates or timestamps. Calendar date when the model was added to OpenRouter.
model_added_year Int64 integer Whole-number numeric column. Calendar year extracted from the OpenRouter added timestamp.
model_age_days Int64 integer Whole-number numeric column. Days between collection time and the OpenRouter added timestamp.
context_length Int64 integer Whole-number numeric column. Primary OpenRouter context window in tokens.
top_provider_context_length Int64 integer Whole-number numeric column. Context window reported by the top provider route.
top_provider_max_completion_tokens Int64 integer Whole-number numeric column. Maximum completion tokens reported by the top provider route.
top_provider_is_moderated bool boolean True/false column. Boolean flag indicating whether OpenRouter marks the top provider as moderated.
architecture_modality object string Text column. Compact architecture string such as text+image+file->text.
input_modalities_pipe object string Text column. Pipe-delimited list of supported input modalities.
output_modalities_pipe object string Text column. Pipe-delimited list of supported output modalities.
input_modality_count Int64 integer Whole-number numeric column. Count of distinct supported input modalities.
output_modality_count Int64 integer Whole-number numeric column. Count of distinct supported output modalities.
tokenizer object string Text column. Tokenizer family label reported by OpenRouter.
instruct_type object string Text column. Instruction-format subtype reported by OpenRouter when available.
primary_market_segment object string Text column. Normalized buyer-facing segment such as text-generation, multimodal-text, embeddings, or image-generation.
context_tier object string Text column. Heuristic context-window bucket derived from the main context length.
supports_multimodal_input bool boolean True/false column. Boolean flag indicating that the model accepts more than plain text input.
accepts_text_input bool boolean True/false column. Boolean flag indicating text input support.
accepts_image_input bool boolean True/false column. Boolean flag indicating image input support.
accepts_audio_input bool boolean True/false column. Boolean flag indicating audio input support.
accepts_video_input bool boolean True/false column. Boolean flag indicating video input support.
accepts_file_input bool boolean True/false column. Boolean flag indicating file input support.
outputs_text bool boolean True/false column. Boolean flag indicating text output support.
outputs_image bool boolean True/false column. Boolean flag indicating image output support.
outputs_audio bool boolean True/false column. Boolean flag indicating audio or speech output support.
outputs_video bool boolean True/false column. Boolean flag indicating video output support.
outputs_embeddings bool boolean True/false column. Boolean flag indicating embedding output support.
outputs_transcription bool boolean True/false column. Boolean flag indicating transcription output support.
outputs_rerank bool boolean True/false column. Boolean flag indicating reranking output support.
knowledge_cutoff object string Text column. Knowledge cutoff date string reported by OpenRouter when available.
knowledge_cutoff_year Int64 integer Whole-number numeric column. Calendar year extracted from the knowledge cutoff string when available.
expiration_date object string Text column. Deprecation or expiration date string reported by OpenRouter when available.
is_deprecated bool boolean True/false column. Boolean flag indicating that OpenRouter currently exposes an expiration date for the model.
supported_parameters_pipe object long text Free-text column containing longer text values. Pipe-delimited list of OpenAI-compatible parameters supported by the model.
supported_parameter_count Int64 integer Whole-number numeric column. Count of supported request parameters exposed by OpenRouter.
supports_tools bool boolean True/false column. Boolean flag indicating function-calling or tools support.
supports_tool_choice bool boolean True/false column. Boolean flag indicating explicit tool-choice parameter support.
supports_reasoning bool boolean True/false column. Boolean flag indicating internal reasoning parameter support.
supports_include_reasoning bool boolean True/false column. Boolean flag indicating include_reasoning parameter support.
supports_structured_outputs bool boolean True/false column. Boolean flag indicating structured_outputs parameter support.
supports_response_format bool boolean True/false column. Boolean flag indicating response_format parameter support.
supports_temperature bool boolean True/false column. Boolean flag indicating temperature parameter support.
supports_top_p bool boolean True/false column. Boolean flag indicating top_p parameter support.
supports_top_k bool boolean True/false column. Boolean flag indicating top_k parameter support.
supports_seed bool boolean True/false column. Boolean flag indicating seed parameter support.
supports_stop bool boolean True/false column. Boolean flag indicating stop parameter support.
supports_frequency_penalty bool boolean True/false column. Boolean flag indicating frequency_penalty parameter support.
supports_presence_penalty bool boolean True/false column. Boolean flag indicating presence_penalty parameter support.
supports_repetition_penalty bool boolean True/false column. Boolean flag indicating repetition_penalty parameter support.
default_parameters_json object string Text column. JSON string of default parameter values exposed by OpenRouter for the model.
prompt_price_per_1m_tokens_usd float64 float Decimal numeric column. Prompt-token price converted to U.S. dollars per one million input tokens.
completion_price_per_1m_tokens_usd float64 float Decimal numeric column. Completion-token price converted to U.S. dollars per one million output tokens.
request_price_usd bool boolean True/false column. Fixed per-request price in U.S. dollars when exposed by OpenRouter.
image_input_price_usd float64 float Decimal numeric column. Per-image price in U.S. dollars when exposed by OpenRouter.
audio_input_price_usd float64 float Decimal numeric column. Per-audio-input price in U.S. dollars when exposed by OpenRouter.
web_search_price_usd float64 float Decimal numeric column. Per web-search operation price in U.S. dollars when exposed by OpenRouter.
internal_reasoning_price_per_1m_tokens_usd float64 float Decimal numeric column. Internal reasoning-token price converted to U.S. dollars per one million tokens when exposed by OpenRouter.
input_cache_read_price_per_1m_tokens_usd float64 float Decimal numeric column. Prompt-cache read price converted to U.S. dollars per one million tokens when exposed by OpenRouter.
input_cache_write_price_per_1m_tokens_usd float64 float Decimal numeric column. Prompt-cache write price converted to U.S. dollars per one million tokens when exposed by OpenRouter.
blended_text_cost_per_1m_in_1m_out_usd float64 float Decimal numeric column. Heuristic combined cost for one million input and one million output text tokens using the published lowest rates.
price_tier object string Text column. Heuristic pricing bucket derived from the blended text cost and free-model status.
is_free_model bool boolean True/false column. Boolean flag indicating that the listed token and request prices are zero or absent.
request_to_prompt_cost_ratio bool boolean True/false column. Per-request price divided by prompt price per one million tokens when both values are available.
endpoint_count Int64 integer Whole-number numeric column. Count of provider endpoints returned by the OpenRouter model-details endpoint.
active_endpoint_count Int64 integer Whole-number numeric column. Count of provider endpoints whose status code is currently active.
provider_count Int64 integer Whole-number numeric column. Count of distinct provider names offering the model through OpenRouter.
provider_diversity_tier object string Text column. Heuristic provider-routing bucket based on the distinct provider count.
provider_names_pipe object string Text column. Pipe-delimited list of provider names serving the model.
provider_tags_pipe object string Text column. Pipe-delimited list of OpenRouter provider tags serving the model.
quantizations_pipe object string Text column. Pipe-delimited list of endpoint quantization labels exposed for the model.
quantization_count Int64 integer Whole-number numeric column. Count of distinct quantization labels exposed across endpoints.
has_region_specific_endpoints bool boolean True/false column. Boolean flag indicating that at least one provider tag includes a region-specific route suffix.
endpoint_context_length_min Int64 integer Whole-number numeric column. Smallest provider endpoint context window exposed for the model.
endpoint_context_length_max Int64 integer Whole-number numeric column. Largest provider endpoint context window exposed for the model.
endpoint_max_completion_tokens_max Int64 integer Whole-number numeric column. Largest provider max-completion-token limit exposed for the model.
endpoint_max_prompt_tokens_max Int64 integer Whole-number numeric column. Largest provider max-prompt-token limit exposed for the model.
best_uptime_last_1d_pct float64 float Decimal numeric column. Highest one-day provider uptime percentage exposed across endpoints when available.
median_uptime_last_1d_pct float64 float Decimal numeric column. Median one-day provider uptime percentage exposed across endpoints when available.
supports_tools_endpoint_count Int64 integer Whole-number numeric column. Count of provider endpoints whose supported-parameters list includes tools.
supports_reasoning_endpoint_count Int64 integer Whole-number numeric column. Count of provider endpoints whose supported-parameters list includes reasoning.
supports_structured_outputs_endpoint_count Int64 integer Whole-number numeric column. Count of provider endpoints whose supported-parameters list includes structured_outputs.
supports_implicit_caching_endpoint_count Int64 integer Whole-number numeric column. Count of provider endpoints that advertise implicit caching support.
per_request_limits_json object string Text column. JSON string of per-request limits metadata exposed by OpenRouter when available.
supported_voices_pipe object long text Free-text column containing longer text values. Pipe-delimited list of supported voices for speech-capable models when available.
source_models_api_url object string Text column. OpenRouter models API URL used for the collection run.
source_model_details_api_url object string Text column. OpenRouter per-model details endpoint used to fetch provider endpoint metadata.
source_docs_url object string Text column. Official OpenRouter models documentation URL reviewed for the source schema.
source_pricing_url object string Text column. Official OpenRouter pricing page reviewed for pricing and plan context.
source_domain object string Text column. Source domain used for the dataset build.
last_collected_at object datetime-like string Stored as text, but values appear to represent dates or timestamps. UTC timestamp when the row was collected for this dataset build.

Intended Use

This dataset is intended for research, experimentation, analysis, and model prototyping.

Loading the Dataset 


import os
from datasets import load_dataset

HUGGINGFACE_API_KEY_KARMANE = os.environ.get("HUGGINGFACE_API_KEY_KARMANE")

dataset = load_dataset(
    "Karmane/openrouter-model-pricing-capability-matrix",
    token=HUGGINGFACE_API_KEY_KARMANE,
)

print(dataset)
print(dataset[list(dataset.keys())[0]][0])

# getting the DataFrame itself
# df = dataset[list(dataset.keys())[0]].to_pandas()

Karmane. (2025). OpenRouter Model Pricing and Capability Matrix. Hugging Face. https://huggingface.co/datasets/Karmane/openrouter-model-pricing-capability-matrix
Downloads last month
5
Total file size:
242 kB