Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding
6 providers
Model catalog
Models
Every model in the catalog with source-linked pricing, context limits, provider availability, and published benchmark results.
Open MoE flagship with million-token context for coding and long agent runs
71 providers
Tencent Hy reasoning model for coding, instruction following, and agent tasks
23 providers
Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work
68 providers
Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy
17 providers
Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads
13 providers
| Model | Creator | Score | Inputs | Context | Input | Output | Released | Compare |
|---|---|---|---|---|---|---|---|---|
| Qwen3.8 Flash Nextalibaba/qwen3.8-flash-next | Alibaba Qwen |
81.0 | 262.144K | $0.1 | $0.2 | 2026-08-27 | ||
| DeepSeek V4 Prodeepseek/deepseek-v4-pro | DeepSeek |
76.2 | 1M | $0.66 | $1.98 | 2026-04-24 | ||
| Hy3tencent/hy3 | Tencent |
75.8 | 256K | $0.066 | $0.26 | 2026-07-06 | ||
| DeepSeek V4 Flashdeepseek/deepseek-v4-flash | DeepSeek |
73.3 | 1M | $0.15 | $0.6 | 2026-04-24 | ||
| Nemotron 3 Ultra 550B A55Bnvidia/nemotron-3-ultra-550b-a55b | NVIDIA |
67.7 | 1M | $0.1 | $0.1 | 2026-06-04 | ||
| Nemotron 3.5 Lightning 30B A3Bnvidia/nemotron-3.5-lightning | NVIDIA |
39.3 | 262.144K | $0.025 | $0.12 | 2026-08-11 |