Cognition's SWE-2 achieves 92.8 on Terminal-Bench 2.1
Cognition releases SWE-2, a 2.8T-parameter MoE coding model post-trained from Kimi K3, scoring 92.8 on Terminal-Bench 2.1.
SWE-2 is a proprietary mixture-of-experts model with 2.8T total parameters and 104B active per token, built on the Kimi K3 base with additional Cognition reinforcement-learning post-training for agentic coding. Vendor-reported benchmarks include FrontierCode 1.1 Main 50.0, DeepSWE 1.1 73.0, Terminal-Bench 2.1 92.8, and Terminal-Bench 4.0 27.3. It claims to be one point behind Claude Fable 5.1 on FrontierCode at a claimed 64% lower cost, but trails Fable 5.1 and GPT-6 Astra by a wide margin on long-horizon Terminal-Bench 4.0 tasks. The model is available today in Devin Desktop and CLI, with no published weights, no per-token API pricing, and all figures pending independent replication.
- SWE-2 is a 2.8T-parameter MoE with 104B active per token, post-trained from Kimi K3
- Self-reported benchmarks: FrontierCode 50.0, DeepSWE 73.0, Terminal-Bench 2.1 92.8, Terminal-Bench 4.0 27.3
- Trails Claude Fable 5.1 and GPT-6 Astra on long-horizon Terminal-Bench 4.0 tasks
- Proprietary weights, no per-token API; available in Devin Desktop and CLI
- Serving stack uses NVFP4/FP8 kernels with quantization-aware training and speculative decoding
- Vendors
- CognitionMoonshot AI
- Organizations
- Cognition
- Countries
- USA
From the original article
2 min read at tokenstead.ai
2. 8T total params, 104B active per token (MoE) - the Kimi K3 base with Cognition’s post-training on top, and the first time Cognition has scaled RL into the multi-trillion-parameter regime. The base had already been RL-heavy for agentic coding; Cognition’s pass added another 5 to 6 points on most benchmarks.
By cdnsteve · tokenstead.ai
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.