Grok-1
xAI's 314B-parameter MoE, released as open weights under Apache 2.0
About
Grok-1 is the raw base-model checkpoint xAI released on March 17, 2024, a 314-billion-parameter mixture-of-experts model with 8 experts and top-2 routing, so roughly 86 billion parameters activate per token. It was the largest open-weight model available at release and xAI's first credibility play with the open-source community. This is the pre-training checkpoint from October 2023, not fine-tuned for chat, and the download is around 300 GB. Third-party assessments at the time placed it roughly alongside Llama 2 70B and Mixtral 8x7B — capable but not frontier. No training data, hyperparameters, or dataset card were published, and running it needs serious GPU memory. Later Grok generations stayed closed.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Third-party assessments at release: roughly competitive with Llama 2 70B and Mixtral 8x7B
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Released 2024-03-17 per xAI announcement and multiple press sources; Apache 2.0 license confirmed; 314B MoE / 8 experts / top-2 routing, raw Oct-2023 base checkpoint (not instruction-tuned); Quality assessment (roughly Llama 2 70B class) is third-party, labeled as such; No training data or hyperparameters published — honest limitation noted.