Baidu logo

ERNIE 5.5 Flash Preview

Baidu

ERNIE 5.5 Flash Preview is Baidu's first post-trained model on the ERNIE 5.5 Flash Base, a mixture-of-experts model with 280B total and 14B active parameters, a 1,048,576-token context window and text-only input and output. Baidu positions it as an intermediate post-training checkpoint strong at coding, 3D modeling and knowledge work, with an official non-preview release planned and Apache 2.0 weights promised. On Qianfan international pay-as-you-go it costs $0.28/$1.13 per million input/output tokens with cache hits at $0.028; thinking is always on and reasoning effort cannot be tuned.

Key Specifications

Parameters
280.0B
Context
1.0M
Release Date
November 11, 2026
Average Score
58.9%

Timeline

Key dates in the model's history
Announcement
November 11, 2026
Last Update
November 12, 2026
Today
November 20, 2026

Technical Specifications

Parameters
280.0B
Training Tokens
-
Knowledge Cutoff
-
Family
-
Capabilities
MultimodalZeroEval

Pricing & Availability

Input (per 1M tokens)
$0.28
Output (per 1M tokens)
$1.13
Max Input Tokens
1.0M
Max Output Tokens
1.0M
Supported Features
Function CallingStructured OutputCode ExecutionWeb SearchBatch InferenceFine-tuning

Benchmark Results

Model performance metrics across various tests and benchmarks

Other Tests

Specialized benchmarks
Agents' Last Exam
Baidu official ERNIE 5.5 Flash Preview launch chart, November 11, 2026. Agents' Last Exam (CLI) 29.0. • Self-reported
29.0%
APEX-Agents
Baidu official ERNIE 5.5 Flash Preview launch chart, November 11, 2026. APEX-Agents 1.1 64.2. • Self-reported
64.2%
AutomationBench
Baidu official ERNIE 5.5 Flash Preview launch chart, November 11, 2026. AutomationBench 61.9. The chart does not name a version, so this uses the generic AutomationBench catalog id. • Self-reported
61.9%
BrowseComp
Baidu official ERNIE 5.5 Flash Preview launch chart, November 11, 2026. BrowseComp 87.4. • Self-reported
87.4%
DeepSWE 1.1
Baidu official ERNIE 5.5 Flash Preview launch chart, November 11, 2026. DeepSWE v1.1 71.2. • Self-reported
71.2%
DRACO
Baidu official ERNIE 5.5 Flash Preview launch chart, November 11, 2026. DRACO 84.9. • Self-reported
84.9%
Humanity's Last Exam (no tools, text-only)
Baidu official ERNIE 5.5 Flash Preview launch table, November 11, 2026. HLE (w/o tools) 46.4; official footnote says this is the text-only subset. • Self-reported
46.4%
MCP Atlas
Baidu official ERNIE 5.5 Flash Preview launch chart, November 11, 2026. MCP-Atlas 84.6. • Self-reported
84.6%
NL2Repo
Baidu official ERNIE 5.5 Flash Preview launch chart, November 11, 2026. NL2Repo 64.8. • Self-reported
64.8%
OfficeQA Pro
Baidu official ERNIE 5.5 Flash Preview launch chart, November 11, 2026. OfficeQA Pro V2 67.9. • Self-reported
67.9%
SpreadsheetBench 2
Baidu official ERNIE 5.5 Flash Preview launch chart, November 11, 2026. SpreadsheetBench V2 36.4. • Self-reported
36.4%
Terminal-Bench 4.0
Baidu official ERNIE 5.5 Flash Preview launch chart, November 11, 2026. Terminal-Bench 4.0 46.5. • Self-reported
46.5%
Terminal-Bench-Science 0.1
Baidu official ERNIE 5.5 Flash Preview launch chart, November 11, 2026. Terminal-Bench-Science 0.1 21.0. • Self-reported
21.0%

License & Metadata

License
proprietary
Announcement Date
November 11, 2026
Last Updated
November 12, 2026

Compare ERNIE 5.5 Flash Preview

All comparisons

Similar Models

All Models

Recommendations are based on similarity of characteristics: developer organization, multimodality, parameter size, and benchmark performance. Choose a model to compare or go to the full catalog to browse all available AI models.