Latest LLM releases & release history

Browse 506 releases, newest first. Current and launch-day Pareto status compare intelligence against benchmark cost. A release qualifies when at least one variant is on the frontier. Unknown dates appear last. Open-weights releases can also lead the open-weights frontier.

506 releases

Showing 409–432 · Page 18 of 22
DeepSeek · 2025-01-20DeepSeek R1 Distill Qwen 14B1 model variantOpen weights
No benchmarks data yet
DeepSeek · 2025-01-20DeepSeek R1 Distill Qwen 32B1 model variantOpen weights
No benchmarks data yet
DeepSeek · 2024-12-26DeepSeek V31 model variantOpen weights
Intelligence
8.50
Total cost
$25.4
Today
Not on frontier
Launch day
Frontier
Google · 2024-12-19Gemini 2.0 Flash Thinking Experimental1 model variant
No benchmarks data yet
OpenAI · 2024-12-17GPT-4o Realtime1 model variant
No benchmarks data yet
OpenAI · 2024-12-17GPT-4o mini Realtime1 model variant
No benchmarks data yet
SpaceXAI · 2024-12-12Grok 21 model variantOpen weights
No benchmarks data yet
Microsoft · 2024-12-12Phi-41 model variantOpen weights
No benchmarks data yet
Google · 2024-12-11Gemini 2.0 Flash1 model variant
No benchmarks data yet
DeepSeek · 2024-12-10DeepSeek-V2.51 model variantOpen weights
No benchmarks data yet
Meta · 2024-12-06Llama 3.3 Instruct 70B1 model variantOpen weights
No benchmarks data yet
OpenAI · 2024-12-05o11 model variant
No benchmarks data yet
Amazon · 2024-12-03Nova Lite1 model variant
No benchmarks data yet
Amazon · 2024-12-03Nova Micro1 model variant
No benchmarks data yet
Amazon · 2024-12-03Nova Pro1 model variant
No benchmarks data yet
Alibaba · 2024-11-27QwQ 32B-Preview1 model variantOpen weights
No benchmarks data yet
Allen Institute for AI · 2024-11-26OLMo 2 7B1 model variantOpen weights
No benchmarks data yet
OpenAI · 2024-11-20GPT-4o1 model variant
No benchmarks data yet
Mistral · 2024-11-18Mistral Large 21 model variantOpen weights
No benchmarks data yet
Mistral · 2024-11-18Pixtral Large1 model variantOpen weights
No benchmarks data yet
Alibaba · 2024-11-18Qwen2.5 Turbo1 model variant
No benchmarks data yet
Alibaba · 2024-11-11Qwen2.5 Coder Instruct 32B1 model variantOpen weights
No benchmarks data yet
Anthropic · 2024-10-22Claude 3.5 Haiku1 model variant
No benchmarks data yet
Anthropic · 2024-10-22Claude 3.5 Sonnet1 model variant
No benchmarks data yet