Model Center
The public BatchIn catalog prioritizes real live models, limited previews, request-access routes, and relay lanes. No filler, no fake cards.
27 listed models
Qwen
qwen3-next-80b-a3b
Featured
Low-cost Qwen route for chat, batch generation, and API consolidation
Price source: SiliconFlow lane
TTFT
160ms
Throughput
120 tok/s
Qwen
qwen3-coder-30b-a3b
Lower-cost Qwen coder lane for tool-using apps, copilots, and internal agents
Price source: SiliconFlow lane
TTFT
160ms
Throughput
120 tok/s
DeepSeek
deepseek-v4-flash
Asia / Batch routes
Fast production DeepSeek route with standard, Asia, and batch pricing lanes
Price source: SiliconFlow lane
TTFT
160ms
Throughput
120 tok/s
Qwen
qwen3.5-397b
Large Qwen route for premium general reasoning with a public batch discount
Price source: SiliconFlow lane
TTFT
420ms
Throughput
58 tok/s
MiniMax
minimax-m2.5
MiniMax pass-through route kept for compatibility and procurement continuity
Price source: SiliconFlow lane
TTFT
160ms
Throughput
120 tok/s
MiniMax
minimax-m2-7
Featured pass-through
Newest MiniMax public lane with transparent pass-through pricing
Price source: SiliconFlow lane
TTFT
160ms
Throughput
120 tok/s
Qwen
qwen-3-6-plus
Higher-end Qwen route offered as a pass-through lane for customers who need latest proprietary capability
Price source: SiliconFlow lane
TTFT
160ms
Throughput
120 tok/s
Moonshot
kimi-k2-6
Featured pass-through
Latest featured Kimi route for customers who need the newest Moonshot capability
Price source: SiliconFlow lane
TTFT
160ms
Throughput
120 tok/s
Moonshot
kimi-k2-5
Previous Kimi production lane with standard and batch pricing
Price source: SiliconFlow lane
TTFT
160ms
Throughput
120 tok/s
StepFun
step-3-5-flash
Fast economical route for broad low-latency text workloads
Price source: SiliconFlow lane
TTFT
160ms
Throughput
120 tok/s
Xiaomi
mimo-v2-flash
Entry MiMo route for cost-sensitive high-volume tasks
Price source: SiliconFlow lane
TTFT
160ms
Throughput
120 tok/s
Xiaomi
mimo-v2-pro
Premium MiMo route for stronger multi-step reasoning and structured responses
TTFT
160ms
Throughput
120 tok/s
OpenAI OSS
gpt-oss-120b
Open-weight GPT-OSS route for low-cost general inference and experimentation
Price source: SiliconFlow lane
TTFT
310ms
Throughput
72 tok/s
Mistral
devstral-2
Developer-focused route for coding assistants and software workflows
Price source: SiliconFlow lane
TTFT
160ms
Throughput
120 tok/s
NVIDIA
nemotron-3-super
NVIDIA route for teams standardizing around enterprise GPU ecosystems
Price source: SiliconFlow lane
TTFT
160ms
Throughput
120 tok/s
Black Forest Labs
flux-schnell
FLUX Schnell is available through BatchIn's public live catalog.
Price source: BatchIn runtime catalog
TTFT
1.4s
Throughput
scene-bound
BAAI
bge-m3
BGE-M3 is available through BatchIn's public live catalog.
Price source: BatchIn runtime catalog
TTFT
160ms
Throughput
120 tok/s
Qwen
qwen3-coder
Qwen3 Coder is available through BatchIn's public live catalog.
Price source: BatchIn runtime catalog
TTFT
220ms
Throughput
94 tok/s
Z.ai
glm-5-1
Featured
Open flagship route for coding, reasoning, and long-horizon agent execution
Price source: SiliconFlow lane
TTFT
520ms
Throughput
42 tok/s
Z.ai
glm-5
High-end GLM lane for production reasoning and long-context workflows
TTFT
520ms
Throughput
42 tok/s
Wan
wan-2.2
Wan 2.2 is available through BatchIn's public live catalog.
TTFT
1.4s
Throughput
scene-bound
Alibaba
cosyvoice
CosyVoice is available through BatchIn's public live catalog.
TTFT
180ms
Throughput
real-time
Mistral
mistral-small-4
Mistral Small 4 is available through BatchIn's public live catalog.
TTFT
160ms
Throughput
120 tok/s
Qwen
qwen3.5-9b
Qwen3.5 9B is available through BatchIn's public live catalog.
TTFT
160ms
Throughput
120 tok/s
Qwen
qwen3-tts
Qwen3 TTS is available through BatchIn's public live catalog.
TTFT
180ms
Throughput
real-time
Meta
llama-4-maverick
Llama 4 Maverick is available through BatchIn's public live catalog.
TTFT
160ms
Throughput
120 tok/s
Qwen
qwen3-coder-next
Qwen3 Coder Next is available through BatchIn's public live catalog.
TTFT
160ms
Throughput
120 tok/s