What's new:
Together AI has updated its model resource management to force cleanup of resources when context managers exit. Additionally, evaluation commands have been removed from the CLI following the service shutdown.
Also shipped Nov 17, 2026Together AI in November 2026 →
Amazon EMR 8.2 is now available
AWS released Amazon EMR 8.2, upgrading to Apache Spark 4.1.3, Apache Iceberg 1.11.0, and Delta Lake 4.2.0 while enabling concurrent Delta Lake writes on Amazon S3.
Introducing the Apache Spark Optimization Agent for Amazon EMR Serverless
AWS introduced the Apache Spark optimization agent for Amazon EMR Serverless to evaluate completed Spark jobs and output automated performance and cost tuning recommendations.
Amazon Bedrock AgentCore Memory now provides full session summaries
Amazon Bedrock AgentCore Memory now automatically creates and maintains a full session summary record for sessions configured with a summary strategy.
[Launched] Generally Available: Intel and AMD v6 with temp disk for Azure Database for PostgreSQL
Microsoft announced general availability of Intel and AMD v6 compute options with local temporary disk for Azure Database for PostgreSQL Flexible Server.
Amazon SageMaker now supports AI data preparation capabilities to build high-quality datasets for model development
Amazon SageMaker introduced AI data preparation features, including synthetic data generation, automated multi-task labeling, human review queues, and quality evaluation.
BigQuery AI.IF and AI.CLASSIFY optimized mode is generally available
Google Cloud made optimized mode generally available for BigQuery managed AI functions AI.IF and AI.CLASSIFY to lower query latency and token usage.
More from Together AI platformEvery Together AI platform release →
New serverless models
Qwen/Qwen4-Max is now available on serverless. It supports a 1,000,000 context length, reasoning, function calling, and structured outputs.
New models available for fine-tuning
NVIDIA-Nemotron-3.5-Lightning-30B-A3B and NVIDIA-Nemotron-3.5-Super-VL-120B-A12B are now available for fine-tuning. Both models support function-calling and reasoning.
What's new:
- Tinker 0.39.0 compatibility: The Tinker-compatible client now builds on tinker 0.39.0, which the tinker extra installs. It follows Tinker 0.39's call styles, so inside a running event loop you can await the same calls you make in synchronous code. ServiceClient now works as a context manager that closes every training client it created.
What's new:
- Rollout capacity budgets: tg endpoints rollout takes two flags that set a rollout's capacity budgets . --max-surge caps how many replicas each canary step adds to the target. --max-unavailable lets a rollout start from a DEGRADED source, or shut down source replicas to make room for the target.
Default expiry for Jig secrets
New Jig secrets now expire after 14 days unless you set a shorter lifetime with --ttl . Before a secret expires, the project's admins and the secret's creator get a daily email reminder. An expired secret is deleted, so new replicas can't read it. Existing replicas keep the value until they terminate.
New serverless models
Together AI launched tencent/Hy4 on its serverless platform, providing 1M context support, NVFP4 quantization, reasoning capabilities, multimodal input, and structured outputs.
Post-training: Nemotron 3.5 Super VL
Together AI added post-training support for NVIDIA Nemotron 3.5 Super VL. The platform supports LoRA training on B200-SXM GPUs with context windows up to 163,840 tokens, rank up to 32, and up to three adapter sessions per resource.
New dedicated endpoint models
The following models are now available for deployment on dedicated endpoints : Qwen/Qwen3-TTS-12Hz-1.7B (text-to-speech). This model was previously listed as Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice . thinkingmachines/Inkling-2-Small .