https://evolution.kevinbytes.com The latest additions to the LLM Tree of Life database, mapping the evolutionary tree of large language models. en-us Sat, 01 Aug 2026 00:00:00 GMT Thu, 20 Aug 2026 10:22:18 GMT https://evolution.kevinbytes.com/models/model-lfm-2.5-vl A fast vision-language model optimized for edge devices, showing strong screen/UI understanding and function-calling. Sat, 01 Aug 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-lfm-2.5-vl community https://evolution.kevinbytes.com/models/model-muse-spark-1.2 Meta's open model tailored for code, launched alongside the 'Muse Code' terminal coding agent. Introduced a 'Contributor' tier for users offering prompt training data. Sat, 01 Aug 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-muse-spark-1.2 meta-llama https://evolution.kevinbytes.com/models/model-qwen3.8-max Alibaba's near-frontier model offering highly competitive performance for its price point. Sat, 01 Aug 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-qwen3.8-max alibaba-qwen https://evolution.kevinbytes.com/models/model-nemotron-3.5-lightning Open model specifically designed for 'always-on' AI agents and specialized business workflows. Sat, 01 Aug 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-nemotron-3.5-lightning nvidia-nemotron https://evolution.kevinbytes.com/models/model-gpt5.6 First frontier model subjected to rigorous US government review prior to launch. Reestablished OpenAI's position at the edge of the cost-performance curve. OpenAI's frontier model family (Sol, Terra, Luna) offering significant improvements. First model to undergo customer-by-customer US government review before launch. Wed, 01 Jul 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-gpt5.6 openai-gpt https://evolution.kevinbytes.com/models/model-claude-opus-5 Delivers near-Fable 5 performance at half the cost. Continues Anthropic's emphasis on high reliability and complex reasoning in the agentic era. Wed, 01 Jul 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-claude-opus-5 anthropic-claude https://evolution.kevinbytes.com/models/model-gemini-3.6-flash A fast, cost-efficient model providing incremental improvements over Gemini 3.5 Flash for high-volume multimodal tasks and low-latency applications. Wed, 01 Jul 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-gemini-3.6-flash google-gemini https://evolution.kevinbytes.com/models/model-gemini-2.5-ultra The most capable variant of the Gemini 2.5 series, showing substantial gains in scientific reasoning and multi-step complex tasks. Wed, 01 Jul 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-gemini-2.5-ultra google-gemini https://evolution.kevinbytes.com/models/model-grok-4.5 Reached Opus-class performance at a highly competitive price point, with tight integrations into SpaceX and Cursor tooling. Wed, 01 Jul 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-grok-4.5 xai-grok https://evolution.kevinbytes.com/models/model-kimi-k3 One of the largest open-weights releases with 2.8 trillion parameters. Wed, 01 Jul 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-kimi-k3 chinese-llm https://evolution.kevinbytes.com/models/model-claude-fable5 First frontier AI model to be suspended by government order, marking a turning point in AI regulation. Demonstrated that the most capable models may face access restrictions beyond the developer's control. First model in Anthropic's 'Mythos-class' tier, representing a new capability level above Opus. Designed for long-horizon agentic work, advanced software engineering, and complex knowledge tasks. Suspended just three days after launch on June 12, 2026, after the U.S. government issued an emergency export-control directive. Mon, 01 Jun 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-claude-fable5 anthropic-claude https://evolution.kevinbytes.com/models/model-claude4.8 Anthropic's current stable frontier model. Features a 1-million-token context window and 128K-token output — capable of processing entire codebases and generating lengthy, detailed responses. Focused improvements on honesty, more efficient tool use, and reliability for professional workflows. Fri, 01 May 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-claude4.8 anthropic-claude https://evolution.kevinbytes.com/models/model-gemini-3.5-flash Google's speed-optimized model designed for high-volume, low-latency applications. Serves as the default model in the Gemini app, balancing strong capabilities with rapid response times and low cost per query. Fri, 01 May 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-gemini-3.5-flash google-gemini https://evolution.kevinbytes.com/models/model-gemini-3.5-pro Google's premium reasoning flagship, unveiled at I/O 2026 with a limited rollout. Represents Google's most capable model for complex coding, analysis, and multi-step reasoning tasks. General availability expected by end of June 2026. Fri, 01 May 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-gemini-3.5-pro google-gemini https://evolution.kevinbytes.com/models/model-grok4 xAI's current flagship model with a 1-million-token context window — enough to process roughly 750,000 words or several full-length novels at once. Features configurable reasoning effort, allowing users to choose between quick responses and deeper, more thorough analysis depending on the task. Available through both the X platform and the X.ai API. Fri, 01 May 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-grok4 xai-grok https://evolution.kevinbytes.com/models/model-command-a-plus Largest fully open-source enterprise model at release, proving that frontier-class multimodal models could be both powerful and completely open. Cohere's most capable model, combining text and vision understanding in a sparse mixture-of-experts design (218B total parameters, 25B active). Supports 48 languages and is fully open-source under Apache 2.0, enabling deployment in government, on-premises, and air-gapped environments. Fri, 01 May 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-command-a-plus cohere-command https://evolution.kevinbytes.com/models/model-qwen3.7 Alibaba's proprietary flagship, available only through their API. Features a massive 1 million token context window (roughly 750,000 words) and is specifically optimized for agentic workflows — tasks where the AI autonomously plans, uses tools, and completes multi-step objectives with minimal human intervention. Fri, 01 May 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-qwen3.7 alibaba-qwen https://evolution.kevinbytes.com/models/model-nemotron3-ultra NVIDIA's answer to GPT-5 and Claude — a 550B MoE monster that uses only 55B active parameters, combining Mamba's linear-time sequence modeling with Transformer attention. NVIDIA's flagship reasoning model and the largest in the Nemotron 3 family. A 550B parameter Mixture-of-Experts model that activates only 55B parameters per computation, combining Mamba's linear-time sequence processing with Transformer attention for both efficiency and quality. Positioned as NVIDIA's answer to frontier models from OpenAI and Anthropic. Fri, 01 May 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-nemotron3-ultra nvidia-nemotron https://evolution.kevinbytes.com/models/model-gpt5.5 OpenAI's current flagship model with a 1 million token context window. Available in Instant (speed-optimized), Thinking (deep reasoning), and Pro (maximum capability) variants. Represents OpenAI's most capable model for complex tasks, creative work, and multi-step agentic workflows. Wed, 01 Apr 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-gpt5.5 openai-gpt https://evolution.kevinbytes.com/models/model-claude4.7 Iterative update focused on reliability — more consistent reasoning, fewer hallucinations (made-up facts), and higher-resolution image understanding. Added new cybersecurity safeguards for responsible use. Wed, 01 Apr 2026 00:00:00 GMT https://evolution.kevinbytes.com/models/model-claude4.7 anthropic-claude