Last verified
COST-EFFICIENT128K CONTEXTTEXTCACHE DISCOUNTCNY SOURCE

ERNIE 4.5 Turbo API Pricing

ERNIE 4.5 Turbo is Baidu's cost-efficient workhorse on Qianfan. The billing page lists CNY 0.8/M input, CNY 3.2/M output, and CNY 0.2/M cache-hit input, which AI//COST stores as $0.119/M, $0.4761/M, and $0.0298/M at 1 CNY = $0.1475. Pulled from cloud.baidu.com.

Input - per 1M tokens
$0.119/M
Original CNY 0.8/M CNY source
Output - per 1M tokens
$0.476/M
Original CNY 3.2/M CNY source
Cached input - 75% off
$0.0298/M
Original CNY 0.2/M -75%
Effective - agentic blend
$0.0803/M
92/8 split - 82% cache
§ 01 / TERMINAL

Run the numbers.

Live calculator pre-loaded with ERNIE 4.5 Turbo's flat per-token rate. The 32K and 128K context variants share the same price on Qianfan.

$ /mo
Workload split
Prompt cache hit rate
Tokens you can process
—
Words equivalent (English)
—
Effective rate
—
Open full calculator (all models · share URL · CSV) →
§ 02 / SCENARIOS

Real-world presets.

CODING AGENT

Codebase-scale migration

$0.010/task
120k cache-heavy in / 10k out~9,831 tasks/$100
LONG DOC ANALYSIS

Reading 100-page contracts

$0.012/doc
180k cache-heavy in / 8k out~8,366 docs/$100
RAG - SUPPORT

Ticket triage

$0.0008/ticket
8k cache-heavy in / 1k out~119,700 tickets/$100
ASSISTANT

Research planning turn

$0.0018/turn
18k cache-heavy in / 2k out~56,761 turns/$100
§ 03 / TOKENIZER

Paste text. See tokens. See cost.

Estimate · ernie-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 521
Words 81
Tokens (estimated) 135 tokens
Cost as input · uncached $0.00002 USD
Cost as output · uncached $0.00006 USD
Cost as cached input $0.000004 USD
§ 04 / SHELF

Up against the shelf.

All models →
Model Input /M Output /M Effective blended Context Best for
ERNIE 4.5 Turbo Current $0.119 cache $0.0298 $0.476 $0.0803 agentic 92/8 128K High-volume cost-sensitive traffic
ERNIE 5.1 $0.595 $2.68 $0.762 frontier sibling 128K Frontier Chinese-language work
DeepSeek V4 Flash $0.44 cache $0.014 $1.32 $0.189 cheaper cross-vendor 1M Bulk low-cost traffic
Qwen 3.5 Flash $0.10 $0.40 $0.124 China peer 1M Cheap long-context Qwen
Gemini 2.5 Flash-Lite $0.10 cache $0.01 $0.40 $0.0561 Western budget peer 1M Cheapest Google multimodal

Frequently asked.

Practical ERNIE 4.5 Turbo pricing questions, with Baidu's published CNY rates kept separate from conversion and workload assumptions.

Q · 01 What is ERNIE 4.5 Turbo priced at? +
On Qianfan it lists CNY 0.8/M input, CNY 3.2/M output, and CNY 0.2/M cache-hit input, stored here as $0.119/M, $0.4761/M, and $0.0298/M at 1 CNY = $0.1475.
Q · 02 How large is the cache discount? +
Cache-hit input is CNY 0.2/M against a base input of CNY 0.8/M - a 75% discount. For cache-heavy agentic workloads, that pulls the effective blended cost down to roughly $0.08/M.
Q · 03 Is the 128K variant more expensive than 32K? +
No. ERNIE 4.5 Turbo has 32K and 128K context variants at the same per-token rate, unlike the ERNIE 5.x models which are tiered by input length.
Q · 04 How does it compare with ERNIE 5.1? +
ERNIE 5.1 is Baidu's frontier tier at $0.5952/$2.6782, roughly 5x the input price of ERNIE 4.5 Turbo. Use Turbo for volume, ERNIE 5.1 for the hardest tasks.
Q · 05 Does this page include tax? +
No. AI//COST stores Baidu's public pre-tax CNY list price and converts to USD for cross-vendor comparison. Enterprise discounts and Baidu AI Cloud resource packages are outside this page.
Y
Reviewed by Yaroslav Vikhariev Founder - AI//COST - ERNIE pricing converted from Qianfan CNY list - Last verified July 11, 2026
Methodology Report a correction More by Y.V.