Last verified
TEXT EMBEDDINGSINPUT ONLYVECTOR INDEXINGNO CACHE DISCOUNTCNY SOURCE

Baichuan-Text-Embedding API Pricing

Baichuan-Text-Embedding is Baichuan's vectorization endpoint for knowledge bases, semantic search, and deduplication rather than chat generation. The official pricing page lists $0.070/M input tokens, converted from 0.0005 yuan per 1K tokens at 7.10 CNY/USD, and there is no separate output-token price. Pulled directly from platform.baichuan-ai.com daily.

Embedding input - per 1M tokens
$0.07/M
Original 0.0005 yuan / 1K input only
Output tokens
$0.00/M
No text output billed $0
Cached input - not listed
$0.07/M
Cache not listed N/A
Effective - embedding blend
$0.07/M
Input-only workload
§ 01 / TERMINAL

Run the numbers.

Live calculator pre-loaded with current Baichuan-Text-Embedding rates. Use token counts from your indexing pipeline to price one-time imports and recurring knowledge-base refreshes.

$ /mo
Workload split
Prompt cache hit rate
Tokens you can process
—
Words equivalent (English)
—
Effective rate
—
§ 02 / SCENARIOS

Real-world presets.

RAG INDEX

Help-center index

$0.000/doc
2k input tokens~710,000 docs/$100
SEARCH

Product snippet embedding

$0.001/item
8k input tokens~177,500 items/$100
DEDUPE

Article dedupe

$0.000/page
600 input tokens~2,366,000 pages/$100
CORPUS

1M-token import

$0.070/corpus
1M input tokens~1,420 imports/$100
§ 03 / TAPE

Price history.

Baichuan's embedding row still sits at $0.070/M input tokens across our verified snapshots.

Input · $0.07/M
Output · $0.00/M
Cached · $0.07/M
JAN 19 Baichuan ended the embedding free beta and started billing at 0.0005 yuan per 1K tokensMAY 23 Live verification kept the same 0.0005 yuan per 1K embedding rate
§ 04 / TOKENIZER

Paste text. See tokens. See cost.

Estimate · baichuan-tokenizer-estimate · ≈3.85 chars/token Auto-counts as you type

This is a chars-per-token approximation, not a real tokenizer. Actual tokens vary by language, code density, and tool-call overhead — counts are typically ±10–20% off for English prose, more for code or non-Latin scripts. For exact billing, use the vendor's official tokenizer.

Characters —
Words —
Tokens (estimated) —
Cost as input · uncached —
Cost as output · uncached —
Cost as cached input —
§ 05 / SHELF

Up against the shelf.

All models →
Model Input /M Output /M Effective blended Context Best for
Baichuan-Text-Embedding Current $0.07 $0.00 $0.07 embedding cost input-only Baichuan-native vector indexing
text-embedding-3-small $0.02 $0.00 $0.02 cheaper embedding peer input-only Low-cost high-volume retrieval
text-embedding-3-large $0.13 $0.00 $0.13 pricier embedding peer input-only Higher-quality OpenAI embeddings
Hunyuan Embedding $0.10 $0.10 $0.10 Tencent embedding peer documented elsewhere Tencent search and retrieval indexing
Baichuan4 Air $0.14 cache $0.14 $0.14 $0.14 text-model budget baseline 32K Lowest-cost Baichuan text traffic

Frequently asked.

Practical Baichuan embedding pricing questions, with the input-only vector workload separated from chat-model assumptions.

Q · 01 What is Baichuan-Text-Embedding priced at today? +
Baichuan's official pricing page lists Baichuan-Text-Embedding at 0.0005 yuan per 1K tokens. AI//COST stores that as $0.0704/M input tokens using the queue's 7.10 CNY/USD conversion rate.
Q · 02 Why does this page show output tokens as $0? +
Because embeddings are an input-only vectorization workload rather than generative chat. The public Baichuan pricing page bills the text you embed and does not publish a separate output-token charge for vectors.
Q · 03 Does Baichuan publish cached-input pricing for embeddings? +
No. The public embedding row does not show a prompt-cache or cache-hit discount, so this page keeps cached input equal to the standard input rate instead of inventing another billing mode.
Q · 04 How does it compare with OpenAI's small embedding model? +
OpenAI's text-embedding-3-small is cheaper at $0.02/M. Baichuan-Text-Embedding costs more on raw input price, but it may still fit better if you want a Baichuan-native Chinese retrieval stack.
Q · 05 Does the pricing page include vector storage? +
No. Baichuan's pricing page separates token billing for the embedding model from file-storage fees in the knowledge-base product. This page only tracks the model token charge, not storage or vector-database costs.
Q · 06 How accurate is the tokenizer estimate? +
The widget uses a baichuan-tokenizer-estimate chars-per-token approximation for English planning. Real embedding bills depend on Baichuan's server-side token count and can differ for Chinese, code, or mixed-language chunks.
Y
Reviewed by Yaroslav Vikhariev Founder - AI//COST - Pricing pulled daily from platform.baichuan-ai.com - Last verified May 23, 2026
Methodology Report a correction More by Y.V.