Model profile
baidu

Baidu: ERNIE 4.5 VL 28B A3B

Baidu: ERNIE 4.5 VL 28B A3B is a budget multimodal generalist from baidu with a heavy runtime profile, compact context posture, and the clearest fit around multimodal / reasoning.

Best for: Multimodal / ReasoningHeavy latencyCompact contextBudget pricing
Intelligence
N/A

Benchmark blend

Coding
N/A

Dev workflow signal

Context
30K Tokens

Compact

Input Price
$0.14

Budget tier

Decision snapshot
48

Baidu: ERNIE 4.5 VL 28B A3B currently reads as a budget multimodal option with compact context and a heavy runtime profile.

Overall profile
Use-case specific
Best for
Multimodal / Reasoning
Latency tier
Heavy
Price tier
Budget
Source coverage
OpenRouterVision signal

Decision Strip

Decision rail before the raw tables

Core buy-side signals stay in one pass. The rest of the page expands only after intelligence, speed, context, and price are clear.

Intelligence
N/A
44

General reasoning and benchmark headroom.

Situational
Speed
N/A
46

Latency data is partial.

Situational
Context
30K Tokens
28

How much prompt and task state can stay in view.

Limited
Price
$0.14
86

$0.56 output / 1M

Efficient

Editorial Profile

Baidu: ERNIE 4.5 VL 28B A3B in one narrative

Positioning, tradeoffs, and fit are consolidated into one read instead of repeating the same story across separate cards.

Use-case specificCoding score 40Math score 36Vision enabled

A powerful multimodal Mixture-of-Experts chat model featuring 28B total parameters with 3B activated per token, delivering exceptional text and vision understanding through its innovative heterogeneous MoE structure with modality-isolated routing. Built with scaling-efficient infrastructure for high-throughput training and inference, the model leverages advanced post-training techniques including SFT, DPO, and UPO for optimized performance, while supporting an impressive 131K context length and RLVR alignment for superior cross-modal reasoning and generation capabilities.

Identity

baidu multimodal profile

Positioning

Multimodal / Reasoning with compact context and heavy runtime.

Cost posture

Efficient spend profile. More comfortable for sustained prompt volume if the capability fit is right.

Strengths
  • Vision-capable routing opens up multimodal review and extraction workflows.

Tradeoffs
  • Budget-friendly input pricing is a strength, but raw capability may vary by workload.

  • Latency profile is better for deliberate runs than rapid back-and-forth chat.

  • Context window is more comfortable for focused tasks than extremely long sessions.

Best fit
  • Image-grounded review, multimodal extraction, and UI audit workflows.

Compare Next

Similar profiles worth opening next

baidu

Baidu: ERNIE 4.5 300B A47B

Intelligence
15.0
Context
123K Tokens
Input Price
$0.28
baidu

Baidu: ERNIE 4.5 21B A3B

Intelligence
N/A
Context
120K Tokens
Input Price
$0.07
baidu

Baidu: ERNIE 4.5 21B A3B Thinking

Intelligence
N/A
Context
131K Tokens
Input Price
$0.07

Benchmarks

Grouped by job-to-be-done

Only benchmark categories with actual signal are shown. Secondary values stay as simple definitions instead of nested micro-cards.

No benchmark data is available for this model yet.

Specs & Pricing

Technical snapshot and cost posture

Specs stay neutral, pricing gets emphasis through values rather than extra containers. Raw provider internals remain in metadata at the end.

Technical snapshot
Context Window
30K Tokens
Vision
Enabled
Modalities
text, image->text, image
Tokenizer
Other
Max Completion
8000
Moderation
No
Supported Parameters
frequency_penaltyinclude_reasoningmax_tokenspresence_penaltyreasoningrepetition_penaltyseedstoptemperaturetool_choicetoolstop_ktop_p
Input Modalities
textimage
Output Modalities
text
Price architecture
Input
per 1M input tokens
$0.14
Output
per 1M output tokens
$0.56
Blended
AA 3:1 mix
N/A

This model is relatively efficient on price. It is the easier fit when sustained prompt volume matters.

Metadata

Raw source tables at the end

Verification details remain available, but the page no longer forces them ahead of the editorial read.