Models / Deep Cogito

Deep Cogito

Deploy Cogito v2 models on Together AI. Iterative self-improvement, 60% shorter reasoning chains, and frontier performance under open license.

Why Deep Cogito on Together AI?

Designed for production workloads that need 
consistent performance and operational control.

Iterative self-improvement

First reasoning models to improve core intelligence, not just search time. Models develop stronger intuition through distillation of reasoning processes — delivering 60% shorter reasoning chains than DeepSeek R1 with superior performance.

Breakthrough efficiency

Complete model family trained for under $3.5M total cost. Significantly more efficient than capital-intensive approaches — proving superintelligence research is accessible to the broader ecosystem.

Open superintelligence

All models released under open license for commercial use. Complete transparency in the reasoning process with visible thinking tags — build on the research or deploy without restrictions.

Meet the Deep Cogito family

Explore top-performing models across text, image, video, code, and voice.

New

Chat

Cogito v2 preview - 671B MoE

New

Chat

Cogito v2.1 671B

New

Chat

Cogito v2 preview - 405B

Chat

Cogito V1 Preview Llama 70B

Chat

Cogito V1 Preview Qwen 32B

Chat

Cogito V1 Preview Qwen 14B

Chat

Cogito V1 Preview Llama 3B

New

Chat

Cogito v2 preview - 109B MoE

New

Chat

Cogito v2 preview - 70B

Breakthrough technical innovations

Explore all the game-changing architectural advances that make Deep Cogito models shine.

  • Mixture of Experts (MoE)

    Sparse expert routing activates only 37B out of 671B parameters for each token in V3. Advanced load balancing without auxiliary losses maintains performance while reducing computational cost.

  • Group Relative Policy Optimization

    New RL approach that removes separate value networks in RLHF, using grouped relative advantage estimation to cut compute requirements while maintaining training stability.

  • Native Reasoning Transparency

    First reasoning model to expose complete thinking process in <think> tags. Native reasoning capabilities built into model foundation through large-scale reinforcement learning.

  • MetaP Training

    First successful implementation of FP8 mixed precision training on a 671B parameter model. Pioneering reinforcement learning approach without supervised fine-tuning as preliminary step.

  • Multi-Head Latent Attention

    Innovative attention mechanism that reduces KV-cache memory requirements while maintaining modeling performance. Optimized for efficient inference deployment.

  • Multi-Token Prediction

    Novel training objective that allows the model to predict multiple tokens simultaneously. Enhanced performance and efficiency through advanced training techniques.

Deployment options

Run models using different deployment options depending on latency needs, traffic patterns, and infrastructure control.

  • Serverless

  • Inference

Serverless Inference

Real-time

A fully managed inference API that automatically scales with request volume.

Best for

Variable or unpredictable traffic

Rapid prototyping and iteration

Cost-sensitive or early-stage production workloads

Batch

Process massive workloads of up to 30 billion tokens asynchronously, at up to 50% less cost.

Best for

Classifying large datasets

Offline summarization

Synthetic data generation

Dedicated Inference

Dedicated Model Inference

An inference endpoint backed by reserved, isolated compute resources and the Together AI inference engine.

Best for

Predictable or steady traffic

Latency-sensitive applications

High-throughput production workloads

Dedicated Container Inference

Run inference with your own engine and model on fully-managed, scalable infrastructure.

Best for

Generative media models

Non-standard runtimes

Custom inference pipelines