AI model library
Google's Gemini 2.5 Flash, released June 2025, is a fast multimodal model accepting text, image, audio, and video input with a 1M-token context window, suited for apps where latency and price matter.
Sol 6.1 is a closed-weight model from OpenAI, part of the sol family. Public information about its capabilities, benchmarks, and pricing was not available at the time of writing.
LFM2.5-24B-A2B is a large language model from Liquid AI's LFM family, released on 2026-10-13 with closed (non-open) weights.
Qwen3-Omni is Alibaba Qwen's omni-modal model that handles text, image, audio and video input and supports speech output, suited to multimodal chat and audio-video understanding.
An embedding model from Perplexity's pplx-embed family, released 2026-10-07. The 'late' variant indicates late-interaction embeddings; weights are closed and access is via API.
Ideogram's image generation model, skilled at creating high-quality artistic images.
NVIDIA's Cosmos 3 model, scheduled for release on September 29, 2026.
xAI's Grok 4.6 is a chat model with text, image and PDF input, designed for long-running agents, coding and visual projects, featuring a 500k token context window.
Daybreak Blue is OpenAI's flagship general-purpose model (gpt-sol family), released 2026-08-07 as proprietary closed weights. It accepts text, image and PDF input, supports chat, reasoning, vision and tool use, with safeguards for defensive cybersecurity work.
Anthropic's Claude Fable 5 model for creative writing, analysis, and controlled agent workflows.
Qwen3.5 27B is an open-weights vision-language model from Alibaba's Qwen team that accepts text, image, and video input. It targets visual reasoning, document understanding, and agent tasks, with a 262k-token context window.