ERNIE 5.1
Baidu·
Overview
MoE flagship derived from ERNIE 5.0 via 'multi-dimensional elastic pre-training' as an optimal sub-network. Parameter counts are estimates (~1/3 of 5.0's total, ~1/2 of 5.0's active params); Baidu has not published exact figures. Cloud-only via Baidu Qianfan, no Western public API. Multimodal (text/image/audio/video) inherited from 5.0. Baidu claims ~6% of comparable training cost — unverified by third parties. No standalone technical report published. Self-reported benchmarks; non-EU hosting.
Capabilities and innovations
Benchmarks
No benchmark values are recorded for this model.
- GPQA Diamond: not reported by the vendor
- Humanity’s Last Exam (no tools): not reported by the vendor
- SWE-bench Verified: not reported by the vendor
- SWE-bench Pro: not reported by the vendor
- MMLU: not reported by the vendor
- MMLU-Pro: not reported by the vendor
- Terminal-Bench 2.x: not reported by the vendor
- MMMU-Pro: not reported by the vendor
API pricing
$0.59 per million input tokens, $2.65 per million output tokens (USD, provider list price, checked 2026-05-11)
Architecture and hardware
- Parameters
- 800B total, 36B active per token (MoE)
Links
More from Baidu
Data curated by AI Model Timeline. See the methodology for admission criteria, benchmark eras and source priority.