unconst/Affine-5czsc2fc98-r35-merged

TEXT GENERATIONPricing:Input $0.4 / Cached $0.07 / Output $4Concurrent Unit Cost:3Model Size:35.1BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Aug 14, 2026Architecture:Transformer Featherless Exclusive Cold

Affine-5czsc2fc98-r35-merged is a 35.1 billion parameter language model developed by unconst, derived from a LoRA-merged checkpoint. This model is specifically noted as a salvage operation from a private TTL insurance, indicating its origin from a fine-tuning process. It features a substantial context length of 32768 tokens, suggesting capabilities for processing extensive inputs. Its primary differentiator lies in its specialized development from a merged checkpoint, likely for a specific, undisclosed application.

Loading preview...

Overview

Affine-5czsc2fc98-r35-merged is a 35.1 billion parameter language model developed by unconst. This model represents a salvaged checkpoint, specifically a LoRA-merged version originating from kevin954/Affine-5dfqbbh8ev-sft. It is described as being part of a private TTL insurance process, indicating a specialized or internal development context rather than a general-purpose release.

Key Characteristics

  • Parameter Count: 35.1 billion parameters.
  • Context Length: Supports a substantial context window of 32768 tokens.
  • Origin: A LoRA-merged checkpoint, salvaged from a previous fine-tuning effort.
  • Development Status: Noted as a private TTL insurance project, awaiting a Stage-5 gate clearance, suggesting it is not yet a public submission.

Potential Use Cases

Given its origin as a salvaged, LoRA-merged checkpoint from a private project, this model is likely intended for:

  • Specialized Internal Applications: Its private nature suggests it's tailored for specific, internal tasks within an organization.
  • Research and Development: Could be used for further experimentation or as a base for more refined models within a controlled environment.
  • Large Context Processing: The 32768-token context length makes it suitable for tasks requiring extensive input analysis.