Model
Qwen3-Next-80B-A3B-Instruct
qwen3-next-80b-a3b-instructRecord
- Family
- qwen
- Architecture
- moe
- Parameters
- 81
- Active parameters
- 3
- HF downloads / 30d
- 6,827
Hugging Face model card
Identity
https://huggingface.co/nvidia/Qwen3-Next-80B-A3B-Instruct-NVFP4- Repository
- nvidia/Qwen3-Next-80B-A3B-Instruct-NVFP4
- Status
- known
- Link type
- Exact Hub repository
Public Hugging Face repository confirmed by the Hub API.
Leaderboard scores
Scraped public quality scores joined to this model by its Hugging Face repository. They describe the listed variant, not any specific local quantization, and are never speed measurements.
| Benchmark | Category | Score | Rank | Scored variant | Confidence |
|---|---|---|---|---|---|
| bfcl | agentic | 65.1 | 32 | Qwen3-Next-80B-A3B-Instruct-FP8 | medium |
| bfcl | agentic | 65.1 | 33 | Qwen3-Next-80B-A3B-Instruct | medium |
| bfcl | agentic | 65.1 | 34 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| bfcl | agentic | 65.1 | 35 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| bfcl | agentic | 65.1 | 36 | Qwen3-Next-80B-A3B-Instruct-bnb-4bit | medium |
| bfcl | agentic | 65.1 | 37 | Qwen3-Next-80B-A3B-Instruct | medium |
| chatbot-arena | chat | 69 | 37 | Qwen3-Next-80B-A3B-Instruct-FP8 | medium |
| chatbot-arena | chat | 69 | 38 | Qwen3-Next-80B-A3B-Instruct | medium |
| chatbot-arena | chat | 69 | 39 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| chatbot-arena | chat | 69 | 40 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| chatbot-arena | chat | 69 | 41 | Qwen3-Next-80B-A3B-Instruct-bnb-4bit | medium |
| chatbot-arena | chat | 69 | 42 | Qwen3-Next-80B-A3B-Instruct | medium |
| livecodebench | coding | 43.2 | 307 | Qwen3-Next-80B-A3B-Instruct-FP8 | medium |
| livecodebench | coding | 43.2 | 308 | Qwen3-Next-80B-A3B-Instruct | medium |
| livecodebench | coding | 43.2 | 309 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| livecodebench | coding | 43.2 | 310 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| livecodebench | coding | 43.2 | 311 | Qwen3-Next-80B-A3B-Instruct-bnb-4bit | medium |
| livecodebench | coding | 43.2 | 312 | Qwen3-Next-80B-A3B-Instruct | medium |
| if-eval | instruction-following | 84.7 | 89 | Qwen3-Next-80B-A3B-Instruct-FP8 | medium |
| if-eval | instruction-following | 84.7 | 90 | Qwen3-Next-80B-A3B-Instruct | medium |
| if-eval | instruction-following | 84.7 | 91 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| if-eval | instruction-following | 84.7 | 92 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| if-eval | instruction-following | 84.7 | 93 | Qwen3-Next-80B-A3B-Instruct-bnb-4bit | medium |
| if-eval | instruction-following | 84.7 | 95 | Qwen3-Next-80B-A3B-Instruct | medium |
| mmlu | knowledge | 89.3 | 38 | Qwen3-Next-80B-A3B-Instruct-FP8 | medium |
| mmlu | knowledge | 89.3 | 39 | Qwen3-Next-80B-A3B-Instruct | medium |
| mmlu | knowledge | 89.3 | 40 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| mmlu | knowledge | 89.3 | 41 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| mmlu | knowledge | 89.3 | 42 | Qwen3-Next-80B-A3B-Instruct-bnb-4bit | medium |
| mmlu | knowledge | 89.3 | 43 | Qwen3-Next-80B-A3B-Instruct | medium |
| mmlu-pro | knowledge | 78.4 | 183 | Qwen3-Next-80B-A3B-Instruct-FP8 | medium |
| mmlu-pro | knowledge | 78.4 | 184 | Qwen3-Next-80B-A3B-Instruct | medium |
| mmlu-pro | knowledge | 78.4 | 185 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| mmlu-pro | knowledge | 78.4 | 186 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| mmlu-pro | knowledge | 78.4 | 187 | Qwen3-Next-80B-A3B-Instruct-bnb-4bit | medium |
| mmlu-pro | knowledge | 78.4 | 188 | Qwen3-Next-80B-A3B-Instruct | medium |
| mmlu-pro | knowledge | 72 | 237 | Qwen3-Next-80B-A3B-Instruct-FP8 | medium |
| mmlu-pro | knowledge | 72 | 238 | Qwen3-Next-80B-A3B-Instruct | medium |
| mmlu-pro | knowledge | 72 | 239 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| mmlu-pro | knowledge | 72 | 240 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| mmlu-pro | knowledge | 72 | 241 | Qwen3-Next-80B-A3B-Instruct-bnb-4bit | medium |
| mmlu-pro | knowledge | 72 | 242 | Qwen3-Next-80B-A3B-Instruct | medium |
| aime-2025 | math | 61.3 | 172 | Qwen3-Next-80B-A3B-Instruct-FP8 | medium |
| aime-2025 | math | 61.3 | 173 | Qwen3-Next-80B-A3B-Instruct | medium |
| aime-2025 | math | 61.3 | 174 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| aime-2025 | math | 61.3 | 175 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| aime-2025 | math | 61.3 | 176 | Qwen3-Next-80B-A3B-Instruct-bnb-4bit | medium |
| aime-2025 | math | 61.3 | 177 | Qwen3-Next-80B-A3B-Instruct | medium |
| gpqa | reasoning | 70.4 | 69 | Qwen3-Next-80B-A3B-Instruct-FP8 | medium |
| gpqa | reasoning | 70.4 | 70 | Qwen3-Next-80B-A3B-Instruct | medium |
| gpqa | reasoning | 70.4 | 71 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| gpqa | reasoning | 70.4 | 72 | Qwen3-Next-80B-A3B-Instruct-GGUF | medium |
| gpqa | reasoning | 70.4 | 73 | Qwen3-Next-80B-A3B-Instruct-bnb-4bit | medium |
| gpqa | reasoning | 70.4 | 74 | Qwen3-Next-80B-A3B-Instruct | medium |
Remaining fields
Identity, launch, related records, and measured speed are shown above. This is the rest of the normalized record.
- active params
- 3
- architecture
- moe
- family
- qwen
- id
- qwen3-next-80b-a3b-instruct
- name
- Qwen3-Next-80B-A3B-Instruct
- params
- 81
- schema version
- local-ai-registry/v1
downloads
- all time
- 263,388
- captured at
- 2026-08-31T12:36:33Z
- last 30d
- 6,827
Provenance & metadata (1)
source huggingface-api
Provenance & metadata (2)
facts
provenance
- captured at
- 2026-08-30T09:26:09Z
sources
| captured at | kind | url |
|---|---|---|
| 2026-08-30T09:26:09Z | normalized-model | huggingface.co/nvidia/Qwen3-Next-80B-A3B-Instruct-NVFP4 ↗ |