Nominate
AI Language Model
Know a product that stood out? Put it forward.
Loading the awards…
← All awards
The language models that made the biggest impact on what AI could do in 2026
9GPT-6 Astra
A reasoning model for complex coding, computer use, research, and professional work, available through OpenAI products and APIs.
Grok 4.6
A Grok model for long-running agents, coding, research, and interactive visual work, with configurable reasoning and a 500,000-token context window.
8Gemini 3.5 Pro
An announced Gemini model that Google reported using internally while preparing a public rollout; current public availability remains unverified.
8DeepSeek-V4
An open-weight language-model family with Pro and Flash variants for reasoning, coding, and agent workflows with million-token context.
8Nano Banana 2
An image-generation and editing model combining Gemini Flash speed with improved text rendering, real-world knowledge, and instruction following.
7Claude Opus 5.5
Anthropic’s latest flagship Claude model for long-running agentic coding and knowledge work. It delivers performance comparable to Claude Fable 5.1 on most tasks at approximately 40% lower cost than Opus 5, with improved safety alignments, reduced jargon, and faster inference. Priced at $4 per million input tokens and $20 per million output tokens.
Mistral Medium 3.5
An open-weight multimodal model optimized for agent and coding workflows, with function calling, structured outputs, and document understanding.
6Inkling
A 975-billion-parameter open-weights multimodal model with controllable reasoning effort, long context, and native text, image, and audio support.
6Kimi K3
An open multimodal model for long-running coding, knowledge work, and reasoning with a one-million-token context window.
5Eleven v4
ElevenLabs' latest text-to-speech model generation, available as Eleven v4 for highest-quality expressive speech and Eleven v4 Turbo for low-latency (~100ms) conversational agents. It improves emotion, voice cloning accuracy from 10-second samples, consistency over long text, and support for 90+ languages compared to v3.
3Jev
TypeSafe AI’s first public System One model, built for fast software automation by returning typed decisions with calibrated probabilities instead of generated text.
3FLUX 3
A multimodal foundation model that generates video with optional synchronized audio and is expanding to image and action prediction.
3Atlas
A multimodal world model in select-partner early access for camera-controlled generation, spatial reconstruction, and simulation from text, images, video, and 3D.
3π0.7
A steerable generalist robot foundation model in development that combines skills and generalizes across unseen tasks, robots, and environments.
0Veo 3.1
A video-generation model creating clips from text and images with synchronized native audio and reference-guided creative controls.