ERNIE-4.5-21B-A3B-Paddle
baidu/ERNIE-4.5-21B-A3B-Paddle
Official · BF16 · safetensors
Runtime certified: Fastdeploy
| Precision / quantization | Format | Bits | Publisher trust | Deployment state |
|---|---|---|---|---|
| BF16 | safetensors | 16 | Official | Runtime certified |
ERNIE · General · ERNIE 4.5
ERNIE 4.5 21B-A3B: Specs, GPU requirements, hosting & deployment
ERNIE 4.5 21B-A3B is a current General release in the ERNIE family, certified from its immutable official source revision.
Total parameters21.0B
Active parameters3.0B
Context window131,072 tokens
Release statusCurrent · Current
ArchitectureErnie4_5_MoeForCausalLM
ModalitiesText
Licenseapache-2.0
Deployment stateDeployable with certified configuration
Commercial stateCommercial Use Allowed
Release date2025-06-28
Architecture, strengths and use cases
ERNIE 4.5 21B-A3B is an official self-hostable Baidu release in the ERNIE General track. It uses Ernie4_5_MoeForCausalLM, carries 21B / 3B MoE, supports a verified context window of 131,072 tokens, and exposes only source-certified checkpoints and runtime combinations. Its release, license and deployment facts are tied to the official repository revision rather than inferred from the family name.
ERNIE 4.5 21B-A3B uses Ernie4_5_MoeForCausalLM with 21B / 3B MoE. The stored architecture profile records 28 layers, hidden width 2560, 20 attention heads and a 131,072 token context ceiling where the official configuration publishes those fields.
Verified Ernie4_5_MoeForCausalLM architecture
131,072 token certified context
Official BF16, Unknown checkpoint coverage
Private LLM inference
Agentic workflows
Long-context document processing
Certified checkpoints and quantizations
Only exact, source-certified variants attached to this release are shown.
Recommended hardware
Use only the listed fastdeploy runtime configuration with the exact checkpoint precision. The recommendation engine checks weight memory, KV cache, runtime overhead, safety headroom, GPU architecture and topology before ranking rentable and reference hardware.
Estimated weight memory40.29 GB
KV cache0.44 GB
Runtime overhead12.05 GB
Total with safety margin65.44 GB
Recommended best value
1× NVIDIA H200 SXM 141GB
141 GB total VRAM · nvswitch
75.56 GB headroom · Limited capacity
USD 2,345 / monthLowest-cost viable
3× NVIDIA L4 24GB
72 GB total VRAM · pcie
6.56 GB headroom · Available
USD 750 / monthHighest performance
1× NVIDIA H200 SXM 141GB
141 GB total VRAM · nvswitch
75.56 GB headroom · Limited capacity
USD 2,345 / monthLicense and self-hosting considerations
The full 21B resident weight set must fit across the selected GPUs. Context and concurrency add KV-cache memory; MoE active parameters do not replace resident-weight memory.
ERNIE 4.5 21B-A3B is published under apache-2.0. The release-scoped official license link and verification date are retained with the catalogue record.
Commercial Use Allowed. The official license permits commercial use; its obligations still apply.
Alternatives and sibling variants
ERNIE 4.5 300B-A47B
Curated comparisons
Release timeline
baidu/ernie/ernie-4-5-21b-a3b
Source details
Source revision: 55376fc68504404a927791015701e9c8b79b9a55d204aa833581c936fab6aeca
Frequently asked questions
How much VRAM does ERNIE 4.5 21B-A3B require?
The exact estimate depends on checkpoint precision, context and concurrency; the page recommendation uses the certified memory engine and compatible runtime configurations.
Which runtimes support ERNIE 4.5 21B-A3B?
Certified upstream runtime evidence is available for fastdeploy.
Can ERNIE 4.5 21B-A3B be used commercially?
The stored commercial classification is Commercial Use Allowed. Consult the official apache-2.0 terms for the final obligations.
Official references
Last verified: 2026-09-03