Terminal-Bench 2.1
Benchmark ● Covered in 12 stories + Follow
Terminal-Bench 2.1 is tracked in TLDRocket’s benchmark dataset. Explore the recorded developments, relationships and supporting evidence below. Coverage reflects our monitored sources and may be incomplete.
Updated 6 October 2026
Latest developments
2026
Model release: Anthropic & Claude Fable 5.1 Model release
Model release: DeepSeek & GLM-5.3 Model release
Open source release: GLM-5.3-Flash & Z.ai Open source release
Model release: Granite 4.2 & Granite Speech 5.0 Turbo CTC Model release
Model release: Meta & Muse Code Model release
Model release: DeepSeek & DeepSeek V4-Flash Model release
Model release: Gemini 3.5 Flash & Gemini Omni Flash Model release
- Market development: Ricursive Intelligence & Ripple-65B Preview
- Market development: Nanbeige LLM Lab & Nanbeige4.3-33B
- Model release: Meta & Muse Spark Open
- Market development: Google Research & Google Cloud AI Research
- Market development: Contrastive-LM & CLM-8B
- Model release: Google & Gemini 3.8 Flash
- Open source release: Z.ai & Hugging Face
- Model release: IBM & Granite 4.2
- Market development: Pokee AI & Pokee-Isaac 28B
- Model release: Meta & Muse Code
- Model release: DeepSeek & DeepSeek V4-Flash-0731
- Model release: Google & Gemini 3.5 Flash
Relationships
Products & technology
- Gemini 3.8 Flash integrated with this benchmark · 1 source
- Granite 4.2 derived from this benchmark · 1 source
- RRSI (Regularized Recursive Self-Improvement) deploys this benchmark · 1 source
- Ripple-65B Preview integrated with this benchmark · 1 source
- Nanbeige4.3-33B integrated with this benchmark · 1 source