Datasets:
File size: 3,488 Bytes
9054080 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 | ---
license: cc-by-4.0
pretty_name: "LLM API prices (OpenAI, Anthropic, Google, DeepSeek)"
language:
- en
tags:
- llm
- pricing
- api
- openai
- anthropic
- google
- deepseek
- tensorust
size_categories:
- n<1K
task_categories:
- tabular-classification
configs:
- config_name: default
data_files:
- split: train
path: prices.csv
---
# LLM API prices: OpenAI, Anthropic, Google, DeepSeek
Per-token API prices for 71 LLM models from 4 providers (67 listed: OpenAI 41, Anthropic 14, Google 10, DeepSeek 2; plus 4 retired Anthropic models kept for history). Every price is copied from the provider's own pricing page, and every row names that page and the date we checked it. Snapshot checked 2026-10-23.
## How it was collected
- Prices: each provider's official pricing page, copied into a dated snapshot (2 checks so far: 2026-10-17 and 2026-10-23).
- Anthropic: https://platform.claude.com/docs/en/about-claude/pricing
- OpenAI: https://developers.openai.com/api/docs/pricing
- Google: https://ai.google.dev/gemini-api/docs/pricing
- DeepSeek: https://api-docs.deepseek.com/quick_start/pricing
- Lifecycle ("current", "legacy", "deprecated", "preview" and so on) and context windows come from each provider's own model pages, checked 2026-10-24.
- No prices are estimated. A blank cell means the provider does not list that price.
## Columns
| column | meaning |
|---|---|
| `provider` | anthropic, openai, google or deepseek |
| `model_id` | the API model id |
| `name` | display name |
| `status` | `listed` on the current pricing page, or `retired` |
| `lifecycle` | provider's own label: current, legacy, deprecated, preview, limited access, shut down, retired, or `no spec page` |
| `input_per_1m`, `output_per_1m` | standard price in USD per 1M input and output tokens |
| `cache_read_per_1m` | USD per 1M cached input tokens read |
| `batch_input_per_1m`, `batch_output_per_1m` | batch API prices in USD per 1M tokens |
| `long_context_threshold` | prompt size in tokens above which the provider lists a separate long-context rate (blank if none) |
| `context_window` | context window in tokens, from the provider's model page |
| `cost_per_1000_chat_replies` | USD for 1,000 chatbot replies of 1,500 input and 400 output tokens at the standard price |
| `source_url` | the pricing page the row was copied from |
| `checked` | date the price was checked |
## Source and updates
Live table, cheapest-model picker and per-model pages: https://tensorust.com/llm/ . Every recorded price change: https://tensorust.com/llm/changes/ . Method: https://tensorust.com/methods/#llm-prices . Latest CSV: https://tensorust.com/llm/prices.csv . The snapshot table and update script are MIT-licensed at github.com/Mattbusel/llm-prices.
Also available through the free Tensorust API (`https://api.tensorust.com/v1/llm/models`, `/v1/llm/cheapest`) and MCP server (`https://mcp.tensorust.com/mcp`).
## Limits
Prices only, no quality scores. Cache-write charges are not in the cost column. Providers change prices; check `checked` and the changelog before quoting a number.
## How to cite
Tensorust Research (2026). *LLM API prices*. https://tensorust.com/llm/
```bibtex
@misc{tensorust2026llmprices,
author = {{Tensorust Research}},
title = {LLM API prices},
year = {2026},
howpublished = {\url{https://tensorust.com/llm/}},
note = {CC BY 4.0}
}
```
|