Qwen · Qwen3 Next
Qwen: Qwen3 Next 80B A3B Thinking
Qwen: Qwen3 Next 80B A3B Thinking is a discovered model from public model catalogues. Badgr builds a GPU serving profile from source metadata, estimated VRAM, context length, and available execution routes.
Last validated or meaningfully updated: Dynamic source discovery
Availability
✓ Badgr AI API
Available through Badgr
✓ Dedicated endpoint
Available through Badgr
✓ Custom GPU deployment
Available through Badgr
Deployment profile
- Architecture
- Source metadata pending
- Licence
- Check model card
- Minimum VRAM
- ~164GB
- Context
- Source dependent
Recommended: H200 141GB. VRAM is an estimate and increases with context, cache, concurrency, and runtime overhead.
GPU deployment scenarios
Estimated from parameter count and quantisation. Will switch to “Verified by Badgr” once a real run backs a tier.
Testing
A100 80GB
Short context, low concurrency
Small production
H100 80GB
8K context, 1-4 concurrent requests
Higher throughput
H200 141GB
32K context, continuous batching
Large production
2× H200 141GB
Tensor parallel, higher concurrency
Estimated — multimodal inputs, long context, and concurrency all increase real VRAM use beyond this estimate.
Also runs well on H100 80GB. Available in United States, Europe.
Serving compatibility
| Runtime | Status | Notes |
|---|---|---|
| vLLM | Estimated | Used for Badgr dedicated endpoints |
| Transformers | Unknown | Basic fallback |
| llama.cpp | Requires GGUF conversion | No GGUF weights found in repo |
| SGLang | Unknown | Not evaluated |
| TensorRT-LLM | Unknown | Conversion may be required |
| TGI | Unknown | Not evaluated |
Call this model through Badgr
curl https://api.aibadgr.com/v1/chat/completions \
-H "Authorization: Bearer $BADGR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"qwen/qwen3-next-80b-a3b-thinking","messages":[{"role":"user","content":"Hello"}]}'Ready-to-run Badgr configurations
Quick test
badgr serve qwen/qwen3-next-80b-a3b-thinking --max-cost 2Dedicated endpoint command
badgr serve qwen/qwen3-next-80b-a3b-thinking --gpu H200 --max-cost 10Advanced configuration (defaults to automatic)
- Runtime
- Quantisation
- GPU and GPU count
- Context length
- Maximum concurrency
- Region
- Maximum hourly spend
- Persistent or capped runtime
Estimated pricing
Provider estimate: cold start
2–5 min
Provider estimate: endpoint cost
$2.50 – $2.50/hr
Idle cost
$0 when stopped
Per-token throughput cost is not shown here: Badgr has not benchmarked this model yet, and this page does not display figures it cannot back with real data.
Recommended Badgr routes
H100 80GB · United States
Availablefrom $2.50/hr
Best for
LoRA Training, vLLM Endpoint, Qwen 2.5 32B Instruct
Startup: 2–5 min
Reliability: High
Last checked: just now
H100 80GB · United States
Availablefrom $2.50/hr
Best for
LoRA Training, vLLM Endpoint, Qwen 2.5 32B Instruct
Startup: 2–5 min
Reliability: High
Last checked: just now
H100 80GB · United States
Availablefrom $2.50/hr
Best for
LoRA Training, vLLM Endpoint, Qwen 2.5 32B Instruct
Startup: 2–5 min
Reliability: High
Last checked: just now
H100 80GB · United States
Availablefrom $2.50/hr
Best for
LoRA Training, vLLM Endpoint, Qwen 2.5 32B Instruct
Startup: 2–5 min
Reliability: High
Last checked: just now
H100 80GB · United States
Availablefrom $2.50/hr
Best for
LoRA Training, vLLM Endpoint, Qwen 2.5 32B Instruct
Startup: 2–5 min
Reliability: High
Last checked: just now
H100 80GB · Europe
Availablefrom $2.50/hr
Best for
LoRA Training, vLLM Endpoint, Qwen 2.5 32B Instruct
Startup: 2–5 min
Reliability: High
Last checked: just now
Known limitations
- Source-discovered pages use best-effort metadata until the model is reviewed.
- Licence, gated access, runner compatibility, and exact memory use should be checked before production deployment.