provider_id stringclasses 7
values | provider stringclasses 7
values | model_id stringlengths 6 22 | model stringlengths 6 22 | input_price_per_1m_tokens float64 0.1 30 | cached_input_price_per_1m_tokens float64 0 1.5 ⌀ | output_price_per_1m_tokens float64 0.28 180 | currency stringclasses 1
value | pricing_unit stringclasses 1
value | status stringclasses 7
values | availability stringlengths 5 38 | official_source_url stringclasses 10
values | last_verified_at stringdate 2026-07-03 00:00:00 2026-08-08 00:00:00 | effective_from stringdate 2024-08-01 00:00:00 2026-07-24 00:00:00 ⌀ | effective_until stringclasses 1
value | notes stringlengths 48 167 ⌀ |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
anthropic | Anthropic | claude-fable-5 | Claude Fable 5 | 10 | 1 | 50 | USD | 1M tokens | latest | Standard | https://platform.claude.com/docs/en/about-claude/pricing | 2026-07-07 | 2026-07-03 | null | Anthropic also lists 5-minute and 1-hour cache write prices. This calculator uses base input, cache hit, and output prices. |
anthropic | Anthropic | claude-haiku-4.5 | Claude Haiku 4.5 | 1 | 0.1 | 5 | USD | 1M tokens | latest | Standard | https://docs.anthropic.com/en/docs/about-claude/pricing | 2026-08-08 | null | null | Lowest-cost current Claude tier in this dataset. |
anthropic | Anthropic | claude-mythos-5 | Claude Mythos 5 | 10 | 1 | 50 | USD | 1M tokens | limited | Limited availability | https://platform.claude.com/docs/en/about-claude/pricing | 2026-07-07 | 2026-07-03 | null | Limited availability. Verify account access before using this in production budgets. |
anthropic | Anthropic | claude-opus-4.1 | Claude Opus 4.1 | 15 | 1.5 | 75 | USD | 1M tokens | legacy | Deprecated | https://platform.claude.com/docs/en/about-claude/pricing | 2026-07-07 | 2026-07-03 | null | Anthropic marks Claude Opus 4.1 as deprecated. Kept only for old forecast comparison. |
anthropic | Anthropic | claude-opus-4.8 | Claude Opus 4.8 | 5 | 0.5 | 25 | USD | 1M tokens | latest | Standard | https://docs.anthropic.com/en/docs/about-claude/pricing | 2026-08-08 | null | null | Anthropic notes Opus 4.7 and later use a new tokenizer that may use more tokens for the same fixed text. |
anthropic | Anthropic | claude-opus-5 | Claude Opus 5 | 5 | 0.5 | 25 | USD | 1M tokens | latest | Standard | https://platform.claude.com/docs/en/about-claude/pricing | 2026-08-08 | 2026-07-24 | null | null |
anthropic | Anthropic | claude-sonnet-4 | Claude Sonnet 4 | 3 | 0.3 | 15 | USD | 1M tokens | retired | Retired except partner clouds | https://platform.claude.com/docs/en/about-claude/pricing | 2026-07-03 | null | null | V2 has no default-safe current price for this row (partially_verified); displayed price is a legacy fallback and must not be treated as V2 verified pricing. |
anthropic | Anthropic | claude-sonnet-4.6 | Claude Sonnet 4.6 | 3 | 0.3 | 15 | USD | 1M tokens | latest | Standard | https://docs.anthropic.com/en/docs/about-claude/pricing | 2026-08-08 | null | null | Good balanced Claude tier for production assistants, coding, and analysis. |
anthropic | Anthropic | claude-sonnet-5 | Claude Sonnet 5 | 2 | 0.2 | 10 | USD | 1M tokens | latest | Standard | https://platform.claude.com/docs/en/about-claude/pricing | 2026-08-08 | 2026-06-30 | 2026-08-31 | Introductory pricing is listed through 2026-08-31: $2 input, $0.20 cache hit, and $10 output per 1M tokens. |
cohere | Cohere | aya-expanse-32b | Aya Expanse 32B | 0.5 | null | 1.5 | USD | 1M tokens | latest | active | https://cohere.com/pricing | 2026-08-08 | null | null | null |
cohere | Cohere | command-r-plus-08-2024 | Command R+ 08-2024 | 2.5 | null | 10 | USD | 1M tokens | latest | active | https://cohere.com/pricing | 2026-08-08 | 2024-08-01 | null | null |
deepseek | DeepSeek | deepseek-v4-flash | DeepSeek V4 Flash | 0.14 | 0.0028 | 0.28 | USD | 1M tokens | latest | Standard | https://api-docs.deepseek.com/quick_start/pricing | 2026-08-08 | null | null | Supports non-thinking and thinking mode. Legacy deepseek-chat and deepseek-reasoner aliases map to this model family until deprecation. |
deepseek | DeepSeek | deepseek-v4-pro | DeepSeek V4 Pro | 0.435 | 0.003625 | 0.87 | USD | 1M tokens | latest | Standard | https://api-docs.deepseek.com/quick_start/pricing | 2026-08-08 | null | null | Higher-capability DeepSeek tier with 1M context and 384K max output. |
google-gemini | Google | gemini-2.5-flash | Gemini 2.5 Flash | 0.3 | 0.03 | 2.5 | USD | 1M tokens | legacy | Legacy reference | https://ai.google.dev/gemini-api/docs/pricing | 2026-08-08 | null | null | Kept for backwards comparison with Gemini 2.5-era forecasts. |
google-gemini | Google | gemini-2.5-pro | Gemini 2.5 Pro | 1.25 | 0.125 | 10 | USD | 1M tokens | legacy | Legacy reference | https://ai.google.dev/gemini-api/docs/pricing | 2026-08-08 | null | null | Kept for teams comparing older Gemini 2.5 forecasts with current Gemini 3.x models. |
google-gemini | Google | gemini-3-flash-preview | Gemini 3 Flash Preview | 0.5 | 0.05 | 3 | USD | 1M tokens | preview | Preview | https://ai.google.dev/gemini-api/docs/pricing | 2026-07-07 | 2026-07-03 | null | Preview model pricing may change before stable release. |
google-gemini | Google | gemini-3.1-flash-lite | Gemini 3.1 Flash-Lite | 0.25 | 0.025 | 1.5 | USD | 1M tokens | active | Standard | https://ai.google.dev/gemini-api/docs/pricing | 2026-07-07 | 2026-07-03 | null | Google also lists Batch, Flex, and Priority pricing for this model. |
google-gemini | Google | gemini-3.1-pro-preview | Gemini 3.1 Pro Preview | 2 | 0.2 | 12 | USD | 1M tokens | preview | Preview | https://ai.google.dev/gemini-api/docs/pricing | 2026-07-07 | 2026-07-03 | null | Prompts over 200K tokens cost more: $4 input, $0.40 cached input, $18 output per 1M tokens. |
google-gemini | Google | gemini-3.5-flash | Gemini 3.5 Flash | 1.5 | 0.15 | 9 | USD | 1M tokens | active | Standard | https://ai.google.dev/gemini-api/docs/pricing | 2026-07-07 | 2026-07-03 | null | Current Gemini Flash tier for higher-quality fast workloads; Google also lists Batch and Priority pricing for eligible usage. |
google-gemini | Google | gemini-3.5-flash-lite | Gemini 3.5 Flash-Lite | 0.3 | 0.03 | 2.5 | USD | 1M tokens | latest | Standard | https://ai.google.dev/gemini-api/docs/pricing | 2026-08-08 | null | null | Generally available since 2026-07-21. Google lists Batch pricing at $0.15 input, $0.02 cached input, and $1.25 output per 1M tokens. |
google-gemini | Google | gemini-3.6-flash | Gemini 3.6 Flash | 1.5 | 0.15 | 7.5 | USD | 1M tokens | latest | Standard | https://ai.google.dev/gemini-api/docs/pricing | 2026-08-08 | null | null | Generally available since 2026-07-21. Google lists Batch pricing at $0.75 input, $0.075 cached input, and $3.75 output per 1M tokens. |
mistral-ai | Mistral AI | mistral-large | Mistral Large 3 | 0.5 | 0.05 | 1.5 | USD | 1M tokens | latest | active | https://mistral.ai/pricing/api/ | 2026-07-07 | null | null | null |
openai | OpenAI | chatgpt-chat-latest | chat-latest | 5 | 0.5 | 30 | USD | 1M tokens | specialized | ChatGPT category | https://developers.openai.com/api/docs/pricing | 2026-07-07 | 2026-07-03 | null | Use this for ChatGPT-style chat-latest estimates only. Production API model IDs may differ. |
openai | OpenAI | gpt-4.1 | GPT-4.1 | 2 | null | 8 | USD | 1M tokens | legacy | Legacy | https://developers.openai.com/api/docs/pricing | 2026-07-03 | null | null | V2 has no default-safe current price for this row (review_required); displayed price is a legacy fallback and must not be treated as V2 verified pricing. |
openai | OpenAI | gpt-4.1-mini | GPT-4.1 mini | 0.4 | null | 1.6 | USD | 1M tokens | legacy | Legacy | https://developers.openai.com/api/docs/pricing | 2026-07-03 | null | null | V2 has no default-safe current price for this row (review_required); displayed price is a legacy fallback and must not be treated as V2 verified pricing. |
openai | OpenAI | gpt-4.1-nano | GPT-4.1 nano | 0.1 | null | 0.4 | USD | 1M tokens | legacy | Legacy | https://developers.openai.com/api/docs/pricing | 2026-07-03 | null | null | V2 has no default-safe current price for this row (review_required); displayed price is a legacy fallback and must not be treated as V2 verified pricing. |
openai | OpenAI | gpt-5.3-codex | GPT-5.3 Codex | 1.75 | 0.175 | 14 | USD | 1M tokens | specialized | Codex | https://developers.openai.com/api/docs/pricing | 2026-07-07 | 2026-07-03 | null | Specialized coding model. Use for code-agent budgeting rather than general chat budgeting. |
openai | OpenAI | gpt-5.4 | GPT-5.4 | 2.5 | 0.25 | 15 | USD | 1M tokens | latest | Standard | https://developers.openai.com/api/docs/pricing | 2026-07-07 | 2026-07-03 | null | Long-context standard pricing is higher: $5 input, $0.50 cached input, $22.50 output per 1M tokens. |
openai | OpenAI | gpt-5.4-mini | GPT-5.4 mini | 0.75 | 0.075 | 4.5 | USD | 1M tokens | latest | Standard | https://developers.openai.com/api/docs/pricing | 2026-07-27 | null | null | Priority pricing is higher. Batch and Flex pricing are lower than Standard for eligible workloads. |
openai | OpenAI | gpt-5.4-nano | GPT-5.4 nano | 0.2 | 0.02 | 1.25 | USD | 1M tokens | latest | Standard | https://developers.openai.com/api/docs/pricing | 2026-07-07 | 2026-07-03 | null | Lowest-cost current OpenAI text tier in this model set; use for routing, extraction, classification, and simple transformations. |
openai | OpenAI | gpt-5.4-pro | GPT-5.4 Pro | 30 | null | 180 | USD | 1M tokens | latest | Premium | https://developers.openai.com/api/docs/pricing | 2026-07-07 | 2026-07-03 | null | Long-context standard pricing is higher: $60 input and $270 output per 1M tokens. Cached input is not listed for this tier. |
openai | OpenAI | gpt-5.5 | GPT-5.5 | 5 | 0.5 | 30 | USD | 1M tokens | latest | Standard | https://developers.openai.com/api/docs/pricing | 2026-07-27 | null | null | Long-context standard pricing is higher: $10 input, $1 cached input, $45 output per 1M tokens. |
openai | OpenAI | gpt-5.5-pro | GPT-5.5 Pro | 30 | null | 180 | USD | 1M tokens | latest | Premium | https://developers.openai.com/api/docs/pricing | 2026-07-07 | 2026-07-03 | null | Long-context standard pricing is higher: $60 input and $270 output per 1M tokens. Cached input is not listed for this tier. |
openai | OpenAI | gpt-5.6-luna | GPT-5.6 Luna | 0.2 | 0.02 | 1.2 | USD | 1M tokens | latest | Standard | https://developers.openai.com/api/docs/pricing | 2026-07-07 | null | null | Long-context standard pricing is higher: $0.40 input, $0.04 cached input, $0.50 cache writes, $1.80 output per 1M tokens. Batch and Flex are lower; Priority is higher. |
openai | OpenAI | gpt-5.6-sol | GPT-5.6 Sol | 5 | 0.5 | 30 | USD | 1M tokens | latest | Standard | https://developers.openai.com/api/docs/pricing | 2026-07-07 | null | null | Long-context standard pricing is higher: $10 input, $1 cached input, $12.50 cache writes, $45 output per 1M tokens. Batch and Flex are lower; Priority is higher. |
openai | OpenAI | gpt-5.6-terra | GPT-5.6 Terra | 2 | 0.2 | 12 | USD | 1M tokens | latest | Standard | https://developers.openai.com/api/docs/pricing | 2026-07-07 | null | null | Long-context standard pricing is higher: $4 input, $0.40 cached input, $5 cache writes, $18 output per 1M tokens. Batch and Flex are lower; Priority is higher. |
xai | xAI | grok-3 | Grok 3 | 1.25 | 0.2 | 2.5 | USD | 1M tokens | retired | Retired; requests redirect to Grok 4.3 | https://docs.x.ai/developers/migration/may-15-retirement | 2026-08-08 | null | null | V2 marks this as a retired historical identity that redirects to Grok 4.3 current billing; it is not an ordinary active recommendation model. |
xai | xAI | grok-4.3 | Grok 4.3 | 1.25 | 0.2 | 2.5 | USD | 1M tokens | active | Standard | https://docs.x.ai/developers/models | 2026-08-08 | null | null | xAI recommends Grok 4.3 for most text use cases and notes model aliases can move to newer versions. |
xai | xAI | grok-4.5 | Grok 4.5 | 2 | 0.3 | 6 | USD | 1M tokens | latest | Standard | https://docs.x.ai/developers/pricing | 2026-08-08 | null | null | Available since 2026-07-16. Above 200K tokens, xAI lists $4 input, $0.60 cached input, and $12 output per 1M tokens. Grok 4.5 is not supported by the Batch API. |
xai | xAI | grok-build-0.1 | Grok Build 0.1 | 1 | null | 2 | USD | 1M tokens | latest | Coding | https://docs.x.ai/developers/models | 2026-07-07 | 2026-07-03 | null | Fast coding model trained for agentic coding workflows. |
AI API Pricing Dataset
This Hugging Face dataset is the machine-readable distribution of the public AI API pricing records published by AICostBudget. It is not a separately curated subset: train.csv, prices.csv, and prices.json are generated from the same Pricing V2 public projection used by the AICostBudget Dataset page and download APIs.
Prices change frequently. Verify production billing decisions against the provider pricing page, contract, billing dashboard, and invoice.
Explore and download
- AICostBudget AI API Pricing Dataset
- Download the public JSON dataset
- Download the public CSV dataset
Turn pricing data into a decision
Calculate your API cost →
Estimate your monthly API bill using these model prices, token usage, and request volume.
Compare models interactively →
Compare OpenAI, Claude, Gemini, DeepSeek, Grok, and other models by API pricing and cost.
Track AI model price changes →
Follow pricing changes that may affect your costs and model choices.
Published files
train.csv: the CSV consumed by the Hugging Face Dataset Viewerprices.csv: the same normalized CSV downloadprices.json: metadata plus the complete public Website record setmeta.json: generation time, verification date, and coverage counts for this export
train.csv and prices.csv are byte-for-byte identical. Unknown or unavailable prices are null in JSON and empty in CSV; they are never rewritten as zero.
Distribution rule
The export is deterministic:
- Start from the checked-in Pricing V2 Website projection.
- Exclude compatibility aliases because they are not independent public pricing rows.
- Exclude rows marked
excluded_default_candidateby the verified pricing pipeline. - Preserve Website legacy fallback rows only when the public Pricing Table adapter still exposes them, including the warning that they are not default-safe verified prices.
- Export exactly the resulting public Website key set; no provider, family, or status-specific allowlist is maintained for Hugging Face.
The canonical provider and pricing facts live in the AICostBudget pricing data repository. Website-owned legacy display metadata is used only for explicitly marked fallback rows. The exporter does not fetch or guess prices.
Fields
| Field | Meaning |
|---|---|
provider_id |
Stable provider identifier. |
provider |
Human-readable provider name. |
model_id |
Stable public model identifier. |
model |
Human-readable model name. |
input_price_per_1m_tokens |
Standard input-token price per 1M tokens. |
cached_input_price_per_1m_tokens |
Cached-input price per 1M tokens when available. |
output_price_per_1m_tokens |
Standard output-token price per 1M tokens. |
currency |
Pricing currency; current records use USD. |
pricing_unit |
Normalized unit; current records use 1M tokens. |
status |
Public lifecycle/status label from the Website projection or marked fallback. |
availability |
Public availability label. |
official_source_url |
Provider source used by the verified pricing pipeline. |
last_verified_at |
Record verification date; it is not refreshed merely because an artifact is rebuilt. |
effective_from |
Provider-stated effective date when available. |
effective_until |
End date when a selected price record has one. |
notes |
Source context, projection warning, or legacy fallback warning. |
Load with pandas
import pandas as pd
url = (
"https://huggingface.co/datasets/"
"aicostbudget-ai/ai-api-pricing/resolve/main/train.csv"
)
pricing = pd.read_csv(url)
print(pricing.head())
print(pricing["provider_id"].value_counts())
Load with Hugging Face Datasets
from datasets import load_dataset
pricing = load_dataset(
"aicostbudget-ai/ai-api-pricing",
split="train",
)
print(pricing)
print(pricing[0])
The train split is the current published pricing snapshot. It is not intended as model-training supervision data.
Direct Hugging Face files
https://huggingface.co/datasets/aicostbudget-ai/ai-api-pricing/resolve/main/train.csv
https://huggingface.co/datasets/aicostbudget-ai/ai-api-pricing/resolve/main/prices.csv
https://huggingface.co/datasets/aicostbudget-ai/ai-api-pricing/resolve/main/prices.json
https://huggingface.co/datasets/aicostbudget-ai/ai-api-pricing/resolve/main/meta.json
Freshness semantics
generated_atis the artifact generation timestamp copied from the checked-in pricing metadata.last_verified_atbelongs to each record and means the price was checked against its source.- Rebuilding this distribution does not rewrite record verification dates.
- Validation requires
generated_at >= max(last_verified_at)and rejects future timestamps.
For current coverage and timestamps, inspect meta.json. The source repository also contains schema checks, price validation, history, snapshots, and freshness workflows.
Methodology and contributions
Prices are accepted from official provider pricing pages, documentation, APIs, or announcements. Third-party aggregators and search snippets are not final price evidence.
Citation
AICostBudget. AICostBudget AI API Pricing Dataset.
https://github.com/aicostbudget/ai-api-pricing-data
Disclaimer
This dataset is informational and may lag provider changes. AICostBudget is independent and is not affiliated with or endorsed by the listed providers. Provider names and trademarks belong to their respective owners.
- Downloads last month
- 516