Releasesby GlyphPost-training: Muse Spark Open
meta-models/Muse-Spark-Open-816B-A42B is available for post-training : LoRA training on B200-SXM , a 159,744-token training context, a 163,840-token generator context, rank up to 32, and one adapter session per resource. See Supported models .
Also shipped Oct 17, 2026Together AI in October 2026 →
Workers, Cloudflare CLI - Test Cron Triggers with Local Explorer
Local Explorer can now send one-off Cron Trigger invocations to a Worker's scheduled() handler during local development. This removes the need to construct request URLs when testing handler behavior. Start your local development server with cf dev , wrangler dev , or the Cloudflare Vite plugin .
AI Gateway logs now show session affinity
AI Gateway logs now have a Sessions view that groups requests into sessions and shows how session affinity routed each one. Each row shows the coding agent or app that ran the session, such as Claude Code, Codex, or Cursor, along with its cost, models, providers, and how much of its input was cached across every turn.
DeepSeek V4.1 Flash 1017 now available on AI Gateway
DeepSeek V4.1 Flash 1017 is now available on AI Gateway . DeepSeek V4.1 Flash 1017 is a minor update to DeepSeek V4.1 Flash that improves reasoning, coding, agentic tasks, and image understanding. The coding gains show up in terminal work and in building and editing whole repositories, and the agentic gains in multi-step automation.
Request a higher workspace role
Workspace viewers and editors can now request a higher role from the People tab. Viewers can ask for editor or admin access, and editors for admin access, each with an optional reason. You can withdraw a request while it is pending. Workspace admins and owners approve or deny it. Admin access requires a paid workspace plan.
Jump to your last chat in the desktop app
In the Lovable desktop app , tap Ctrl+Tab to switch to the chat you were in before. Hold Ctrl and press Tab to pick from your recent chats instead, with the chat you are in first. The shortcut uses Ctrl on macOS as well.
More from Together AI platformEvery Together AI platform release →
New serverless models
The following models are now available on serverless : deepseek-ai/DeepSeek-V4.1-Flash-1017 : 1,000,000 context length, FP8 quantization, function calling, and structured outputs. Pricing: $0.30 input / $1.20 output / $0.006 cached input (per 1M tokens).
New dedicated endpoint models
The following models are now available for deployment on dedicated endpoints : deepseek-ai/DeepSeek-V4.1-Flash-1017 .
New serverless models
The following models are now available on serverless : reflection/Beam-1.0-501B-A23B : 1,048,576 context length, NVFP4 quantization, function calling, and structured outputs. Pricing: $0.40 input / $1.20 output / $0.04 cached input (per 1M tokens).
Geo-specific inference
You can now keep serverless inference processing in the United States or Canada. Set your client's base URL to https://us-api.together.ai/v1 for the United States or https://ca-api.together.ai/v1 for Canada, using your existing API key and model IDs. There are no GPUs to provision.
New dedicated endpoint models
The following models are now available for deployment on dedicated endpoints : moonshotai/Kimi-K3 .
Longer post-training context for Kimi K3
moonshotai/Kimi-K3 on post-training now supports a 163,840-token training context (up from 65,536) and a 167,936-token generator context (up from 69,632). See Supported models .
New serverless models
The following models are now available on serverless : meta-models/Muse-Spark-Open-816B-A42B : 1,048,576 context length, NVFP4 quantization, function calling, and structured outputs. Pricing: $1.25 input / $4.25 output / $0.15 cached input (per 1M tokens).
Post-training
Post-training lets you write your own training loop and run it against Together AI GPUs. Your code builds each batch and chooses the loss function and learning rate, while the GPUs you provision run the forward and backward passes, optimizer steps, and sampling. It bills per GPU-minute.