Total 1,098
On 24 April 2026 DeepSeek released the DeepSeek-V4 preview, two open mixture-of-experts models with a 1M-token context that became the default across its official services. Both models support thinking and non-thinking modes.
OpenAI released GPT-5.5 Thinking and GPT-5.5 Pro to paid ChatGPT users on 23 April 2026, then extended the models to Codex and the API, reporting strong benchmark results and later making a faster GPT-5.5 Instant the default.
On 23 April 2026 Tencent's Hy Team open-sourced the Hy3 preview on Hugging Face, ModelScope and GitCode, a 295-billion-parameter mixture-of-experts model with a 256K context.
On 21 April 2026 OpenAI released gpt-image-2, an image model that generates and edits pictures from text and image inputs and appears in ChatGPT as ChatGPT Images 2.0. It replaced the GPT Image 1 line and was later revised as versions 2.5 and 2.6.
On 17 April 2026, Anthropic Labs launched Claude Design, a Claude Opus 4.7-powered research preview for Pro, Max, Team and Enterprise subscribers to create designs, prototypes, slides and one-pagers.
On 16 April 2026 Anthropic made Claude Opus 4.7 generally available at $5 and $25 per million tokens. It was less broadly capable than Claude Mythos Preview, with deliberately reduced cyber capabilities, and was the first model to test automatic cyber safeguards.
Shengshu Technology announced a Series B round of nearly 2 billion yuan led by Alibaba Cloud, money it says will go to foundation world models for both digital generation and physical action.
On 9 April 2026 LG AI Research released EXAONE 4.5, its first open-weight multimodal model: a 33B vision-language system aimed at document understanding, with a 256K context window and the earlier EXAONE licence.
On 8 April 2026 HeyGen announced Avatar V, which it describes as the most realistic AI avatar model in the world; a recording of about 15 seconds is enough to generate video that keeps a person's face, voice and presence.
On 8 April 2026 Meta Superintelligence Labs released Muse Spark, the first model from the lab and the first of the Muse series. It is closed: its weights were not released, and it was offered through a private API preview.
Zhipu released GLM-5.1 to its subscribers in late March 2026 and open-sourced it under the MIT License on 7 April 2026. Its model card reports 58.4 on SWE-Bench Pro, the top score in its comparison table.
On 7 April 2026 Anthropic announced Project Glasswing, which gave a restricted group access to the unreleased Claude Mythos Preview to help secure critical software.
On 2 April 2026 Google DeepMind released Gemma 4, a family of open models ranging from small edge versions to a 31B dense model, now published under the commercially permissive Apache 2.0 license.
In April 2026 Alibaba launched Qwen3.6, starting with the proprietary Qwen3.6-Plus for agentic coding on 2 April, followed by open Qwen3.6-35B-A3B weights in mid-April.
On 2 April 2026 Mustafa Suleyman announced three Microsoft AI models for developers: MAI-Transcribe-1, MAI-Voice-1 and MAI-Image-2. All three became available in Microsoft Foundry and the MAI Playground, with published model cards.
In April 2026 Pony.ai, with Verne and Uber, launched a commercial robotaxi service in Zagreb, Croatia, which the company describes as Europe's first. It was part of a China-plus-overseas strategy that also included Singapore and Luxembourg testing.
In April 2026 Moonshot AI released Kimi K2.6, an open-weight model aimed at long-horizon coding, with Agent Swarm scaled to 300 sub-agents and 4,000 coordinated steps.
On 24 March 2026 OpenAI announced that it was discontinuing Sora in its app and API. The Sora app shut down on 26 April 2026, and the planned Disney investment tied to Sora did not go ahead.
On 23 March 2026 Mistral AI released Voxtral TTS, a 4B text-to-speech model with voice cloning from three seconds of audio, open weights under a non-commercial licence, and API pricing of $0.016 per 1,000 characters.
Midjourney released the V8 alpha on 17 March 2026, generating images about five times faster, with a native 2K --hd mode, improved text rendering, new prompt parameters and a Grid Mode interface. Point releases followed later that year.
On 16 March 2026, Mistral AI released Mistral Small 4, a 119B mixture-of-experts model under Apache 2.0 that combines reasoning, image input and agentic coding in one model.
