Trilogix1/OmniCoder-9B

VISIONPricing:Input $0.431 / Cached $0.0862 / Output $1.12Concurrent Unit Cost:1Model Size:9BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Oct 14, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

OmniCoder-9B is a 9-billion parameter coding agent model developed by Tesslate, built on Qwen3.5-9B's hybrid architecture. It was fine-tuned on over 425,000 agentic coding trajectories, including real-world software engineering tasks and tool use. This model excels at complex coding problems, error recovery, and multi-step reasoning, making it ideal for advanced code generation and agentic applications.

Loading preview...

OmniCoder-9B: A 9B Coding Agent

OmniCoder-9B, developed by Tesslate, is a 9-billion parameter coding agent model based on the Qwen3.5-9B's hybrid architecture, which incorporates Gated Delta Networks for efficient long-context processing. This model is uniquely fine-tuned on over 425,000 curated agentic coding trajectories, derived from advanced models like Claude Opus 4.6, GPT-5.4, GPT-5.3-Codex, and Gemini 3.1 Pro. It features a 262K native context window, extensible to over 1M tokens.

Key Capabilities

  • Advanced Agentic Behavior: Learns from frontier agent traces, demonstrating strong error recovery (read-before-write), responsiveness to LSP diagnostics, and precise edit diffs.
  • Complex Problem Solving: Supports <think>...</think> reasoning chains for decomposing and solving intricate coding challenges.
  • High Performance: Achieves notable scores on coding and reasoning benchmarks, including 83.8% on GPQA Diamond (pass@1) and 23.6% on Terminal-Bench 2.0, significantly outperforming its base model.
  • Open Weights: Released under the Apache 2.0 license, offering full flexibility for development and deployment.

Good For

  • Automated Software Engineering: Ideal for tasks requiring multi-step reasoning, tool use, and robust error handling in coding environments.
  • Code Generation & Refactoring: Excels at generating high-quality code and applying minimal, precise edits.
  • Agentic Applications: Suitable for building intelligent coding assistants and autonomous development agents that can learn from and adapt to complex scenarios.
  • Research & Development: Provides a strong foundation for exploring advanced agentic AI and coding models due to its open-source nature and unique training methodology.

Finetunes
1 models