Together AI Launches Kimi K3 API, Leads Key Inference Benchmarks
togethercompute · x · 2026-08-06
Together AI has launched the API for Moonshot AI's Kimi K3 model. Kimi K3 is the first open-source model in the 3-trillion-parameter class, featuring 2.8T total parameters with a sparse MoE architecture that activates 16B per token. It supports native vision and a 1-million-token context window.
According to benchmarks by Moonshot, Together AI ranks #1 or ties for #1 on 3 out of 4 major benchmarks (OCRBench, MMMU Pro Vision, and DeepSWE). Together AI emphasized that as open models like Kimi K3 achieve frontier-class quality, serving them with high-performance infrastructure becomes essential.
More from Infra
- Open-Source Benchmarks: RTX 5090 LLM Quants and 8GB VRAM Agentic Scores — max_paperclips · 2026-08-06
- NVIDIA Discusses Building Secure Enterprise AI with Proprietary Data — nvidia · 2026-08-06
- Chorus: Open-Source Pre-trained Model Library Enables Fast CPU Inference Without GPUs — jmschreiber91 · 2026-08-06
- Running DeepSeek V4 Locally on Spark Hardware Hits ~95 tok/s — Rasmic · 2026-08-06
- Local Deployment: Running an NVIDIA and AMD GPU Together for Different Models — Curious-Pen5547 · 2026-08-06
- Luminal Compiler Discovers Insanely Fast Megakernels Without Quantization — AccBalanced · 2026-08-06