Kurrens serves open-weight models on infrastructure we operate. Every model is identified by its Hugging Face id and lists the numerical precision we serve it at. Click an id to copy it.
21 of 21 models
- Context
- 1M
- Max output
- 922K
- Precision
- FP8
- License
- MIT
- Context
- 1M
- Max output
- 384K
- Precision
- FP8
- License
- MIT
- Context
- 1.3M
- Max output
- 642K
- Precision
- FP8
- License
- MIT
- Context
- 1.3M
- Max output
- 256K
- Precision
- FP8
- License
- GLM-5.3 License
- Context
- 1.3M
- Max output
- 243K
- Precision
- FP8
- License
- MIT
- Context
- 1M
- Max output
- 230K
- Precision
- FP8
- License
- MIT
- Context
- 1M
- Max output
- 922K
- Precision
- FP8
- License
- Kimi K3 License
- Context
- 256K
- Max output
- 230K
- Precision
- FP8
- License
- Modified MIT
- Context
- 256K
- Max output
- 230K
- Precision
- FP8
- License
- Modified MIT
- Context
- 1M
- Max output
- 500K
- Precision
- FP8
- License
- MiniMax Community License
- Context
- 200K
- Max output
- 128K
- Precision
- FP8
- License
- Custom license
- Context
- 1M
- Max output
- 230K
- Precision
- FP8
- License
- Apache 2.0
- Context
- 1M
- Max output
- 128K
- Precision
- FP8
- License
- Qwen Community License 1.0
- Context
- 1M
- Max output
- 128K
- Precision
- FP8
- License
- Qwen3.8 Max License
- Context
- 256K
- Max output
- 179K
- Precision
- FP8
- License
- Apache 2.0
- Context
- 256K
- Max output
- 64K
- Precision
- FP8
- License
- Apache 2.0
- Context
- 1M
- Max output
- 922K
- Precision
- FP8
- License
- MIT
- Context
- 1M
- Max output
- 128K
- Precision
- FP8
- License
- MIT
- Context
- 128K
- Max output
- 115K
- Precision
- FP8
- License
- Apache 2.0
- Context
- 128K
- Max output
- 115K
- Precision
- FP8
- License
- Apache 2.0
- Context
- 256K
- Max output
- 64K
- Precision
- FP8
- License
- Apache 2.0
No models match.
Prices are listed on kurrens.ai/pricing. During pre-launch they are estimates and may change.
Listing models programmatically
curl https://api.kurrens.ai/v1/models -H "Authorization: Bearer $KURRENS_API_KEY"A public, unauthenticated copy of the catalog is published at kurrens.ai/models.json in the OpenRouter provider schema (v2.4).
Precision and versions
A model id always points to the same weights at the same precision. New releases get a new id. See Precision & versioning.
Licenses
Each model is released by its developer under its own license. Your use of a model is subject to that license — see Model Licenses.
Retirements
When a model is scheduled for retirement we announce it at least 30 days in advance. After the retirement date, requests
return 404 model_retired. See Model lifecycle and the Changelog.