Prime Intellect
Prime Intellect operates an AI training platform and research lab. Its own releases include the INTELLECT series and small models used to study reinforcement learning.
Models & tools
- INTELLECT-1
Prime Intellect · INTELLECT
A 10-billion-parameter base language model trained across distributed computers.
Launch date Downloadable - glm4-moe-tiny
Prime Intellect · Prime RL research models
A small GLM-style mixture-of-experts model for training-system research.
Launch date not confirmedResearch releaseDownloadable - INTELLECT-1-Instruct
Prime Intellect · INTELLECT
The instruction-trained version of the 10-billion-parameter INTELLECT-1 model.
Launch date not confirmedDownloadable - INTELLECT-2
Prime Intellect · INTELLECT
A 32-billion-parameter reasoning model trained with distributed reinforcement learning.
Launch date not confirmedDownloadable - INTELLECT-3
Prime Intellect · INTELLECT
A 106-billion-parameter mixture-of-experts reasoning model with 12 billion active parameters.
Launch date not confirmedDownloadable - INTELLECT-3.1
Prime Intellect · INTELLECT
A further reinforcement-learning release of the INTELLECT-3 reasoning model.
Launch date not confirmedDownloadable - INTELLECT-MATH
Prime Intellect · INTELLECT-MATH
A seven-billion-parameter math model trained with supervised learning and reinforcement learning.
Launch date not confirmedResearch releaseDownloadable - INTELLECT-MATH-SFT
Prime Intellect · INTELLECT-MATH
A supervised math-reasoning model that supplies the starting policy for INTELLECT-MATH.
Launch date not confirmedResearch releaseDownloadable - Qwen3-0.6B-Reverse-Text-RL
Prime Intellect · Prime RL research models
The reinforcement-learning stage of Prime's Qwen3 text-reversal example.
Launch date not confirmedResearch releaseDownloadable - Qwen3-0.6B-Reverse-Text-SFT
Prime Intellect · Prime RL research models
A small Qwen3 fine-tune that learns to reverse text.
Launch date not confirmedResearch releaseDownloadable - Qwen3-1.7B-Wordle-RL
Prime Intellect · Prime RL research models
A Qwen3 model trained to improve Wordle play through multi-turn rewards.
Launch date not confirmedResearch releaseDownloadable - Qwen3-1.7B-Wordle-SFT
Prime Intellect · Prime RL research models
A supervised Qwen3 fine-tune that learns the response format for Wordle.
Launch date not confirmedResearch releaseDownloadable - Qwen3-4B-Instruct-AlphabetSort-RL
Prime Intellect · Prime RL research models
A Qwen3 fine-tune that maintains alphabetically sorted names across turns.
Launch date not confirmedResearch releaseDownloadable - qwen3-moe-tiny
Prime Intellect · Prime RL research models
A small Qwen3-style mixture-of-experts model for training-system research.
Launch date not confirmedResearch releaseDownloadable - Qwen3.5-0.8B-Reverse-Text-RL
Prime Intellect · Prime RL research models
A short reinforcement-learning run for Qwen3.5 training-system checks.
Launch date not confirmedResearch releaseDownloadable - Qwen3.5-0.8B-Reverse-Text-SFT
Prime Intellect · Prime RL research models
A deliberately undertrained Qwen3.5 model for training-system research.
Launch date not confirmedResearch releaseDownloadable - SYNTHETIC-1-SFT-7B
Prime Intellect · SYNTHETIC-1
A Qwen2.5-based reasoning model trained on collaboratively generated examples.
Launch date not confirmedResearch releaseDownloadable