Large language model
Verdict
DeepSeek V4 Pro 0423 from DeepSeek. Open weights (MIT). 1.0M context. $0.9553/$1.9105 per 1M tokens. Artificial Analysis Intelligence 31. Arena text #55.
Training data policy: N/A — open weights. Self-host for full control.
DPDP Act (India): Self-host in India for full DPDP compliance.
Data residency: Self-hosted
Self-host in India