Nebius Token Factory models shut down on November 9, 2026
Nebius Token Factory shut down 11 models on November 9, 2026: deepseek-ai/DeepSeek-V4-Flash-0731, deepseek-ai/DeepSeek-V4-Pro, deepseek-ai/DeepSeek-V4.1-Flash, moonshotai/Kimi-K2.6, moonshotai/Kimi-K2.7-Code, nvidia/nemotron-3-super-120b-a12b, nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B, openbmb/MiniCPM-V-4_5, Qwen/Qwen3-235B-A22B-Instruct-2507, Qwen/Qwen3.5-397B-A17B, and zai-org/GLM-5.1. Official replacements named by the provider: deepseek-ai/DeepSeek-V4-Pro-0813, deepseek-ai/DeepSeek-V4.1-Flash-1017, moonshotai/Kimi-K3, nvidia/Nemotron-3_5-Lightning, nvidia/Nemotron-3.5-Super-VL, and Qwen/Qwen3-30B-A3B-Instruct-2507, plus 2 more. Requests to these models started failing on that date.
Find every request that still uses these model ids. They shut down on November 9, 2026.
Point each request at the replacement in the table. If a cell says "Not published", the provider did not name one.
Open the source link on each row and read that page before you switch.
Run the scan page on your code to catch the ids in this table.
Scan your code for these model ids. Each migrate link opens the diff for that model.
SunsetDiff Pro is in early access. Join free today. You won't be charged until Pro launches, and we'll email you before your first charge. After that it's $19 a month, and you can cancel anytime. Alerts only for the models in your code are coming at launch. The form on this page is the free alert signup.
Questions
Which Nebius Token Factory models shut down on November 9, 2026?
Nebius Token Factory shut down these 11 models on November 9, 2026: deepseek-ai/DeepSeek-V4-Flash-0731, deepseek-ai/DeepSeek-V4-Pro, deepseek-ai/DeepSeek-V4.1-Flash, moonshotai/Kimi-K2.6, moonshotai/Kimi-K2.7-Code, nvidia/nemotron-3-super-120b-a12b, nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B, openbmb/MiniCPM-V-4_5, Qwen/Qwen3-235B-A22B-Instruct-2507, Qwen/Qwen3.5-397B-A17B, and zai-org/GLM-5.1.
What replaces the Nebius Token Factory models that shut down on November 9, 2026?
Official replacements named by the provider: deepseek-ai/DeepSeek-V4-Pro-0813, deepseek-ai/DeepSeek-V4.1-Flash-1017, moonshotai/Kimi-K3, nvidia/Nemotron-3_5-Lightning, nvidia/Nemotron-3.5-Super-VL, and Qwen/Qwen3-30B-A3B-Instruct-2507, plus 2 more. The table lists the replacement for each model.
On what date did requests to these models fail?
These models shut down on November 9, 2026. Requests started failing on that date.