AI model release history.

239 major releases from OpenAI, Anthropic, Google, DeepSeek, Alibaba, Z.ai and Moonshot AI — 142 of them new models — dated from the makers’ own announcements. The newest model recorded here is Images 2.6 (4 Nov 2026).

239 MAJOR RELEASES7 LABSCHECKED 6 NOV 2026
NEWEST FROM OPENAI
Images 2.6 · 4 Nov 2026
NEWEST FROM ANTHROPIC
Claude Haiku 5.5 · 7 Oct 2026
NEWEST FROM GOOGLE
Gemini 4 Carbon · 28 Oct 2026
NEWEST FROM DEEPSEEK
DeepSeek-V4.1-Flash · 10 Sep 2026
NEWEST FROM ALIBABA
Qwen3.8-Flash-Next · 26 Aug 2026
NEWEST FROM Z.AI
GLM-5.3-Flash · 26 Aug 2026
NEWEST FROM MOONSHOT AI
Kimi K3.1 Pro · 24 Oct 2026
§ 01 — LATEST

The latest major LLM and AI model releases.

ACROSS ALL SEVEN LABSNEWEST FIRST
Personal Finance in Claude5 NOV 2026 · ANTHROPICAnthropic introduced Personal Finance, a hub in Claude where users connect bank and card accounts through Plaid with read-only access and ask about balances, spending, bills and fees. It cannot move or invest money, and hub conversations aren’t used to train Anthropic’s models.ProductImages 2.64 NOV 2026 · OPENAIOpenAI launched Images 2.6 in ChatGPT, Codex and the API, generating images 50% faster than before and adding animated GIFs. In the API it released GPT-Image-2.6 Flare and GPT-Image-2.6 Sunburst, with SVG output, markup-guided editing and a new complexity setting.ModelGemini 4 Carbon28 OCT 2026 · GOOGLEGoogle announced Gemini 4 Carbon, its strongest model, for long-horizon engineering and cybersecurity defence, and began rolling it out to a subset of cyber defenders through its Fairwind Program while it takes part in U.S. government pre-release testing. Google says it will launch at Argon’s introductory price of US$2 per million input tokens and US$10 per million output tokens.ModelGPT-6.1 Terra26 OCT 2026 · OPENAIOpenAI released GPT-6.1 Terra, a fast, efficient model for visual reasoning, computer use and writing, in ChatGPT Work, Codex and the API. Standard API pricing is US$1 per million input tokens and US$5 per million output tokens, half the price of GPT-6.1 Sol, with a 1,050,000-token context window.ModelKimi K3.1 Pro24 OCT 2026 · MOONSHOT AIKimi K3.1 Pro, the first model in the K3.1 series, is a post-trained version of Kimi K3 with a 1-million-token context, released with a Highspeed variant on Kimi.ai, the apps, Kimi Work, Kimi Code and the API. Full weights followed on Hugging Face in early November 2026 under the Kimi K3.1 Pro licence.ModelGemini 4 Argon GA13 OCT 2026 · GOOGLEGemini 4 Argon became globally available to developers and enterprises through the Gemini API in AI Studio and the Gemini platform, to Google AI Ultra subscribers in Antigravity, and to AI Ultra and Pro subscribers in the Gemini app and AI Mode in Search. The introductory API price is US$2 per million input tokens and US$10 per million output tokens until 30 November 2026, then US$4 and US$20.ModelGPT-6 and Intelligent UI in ChatGPT7 OCT 2026 · OPENAIOpenAI brought GPT-6 to the Chat tab for every ChatGPT tier, powered by GPT-6 Sol on paid plans and GPT-6 Luna on Free and Go. It introduced Intelligent UI, where answers can include charts, maps, forms, buttons and small interactive tools, and lets ChatGPT start answering while it continues to think.ProductClaude Haiku 5.57 OCT 2026 · ANTHROPICAnthropic released Claude Haiku 5.5 for high-volume, cost-sensitive work such as summaries, classification, subagents and browser use, with a 1M-token context window, 128K-token output limit and the first adjustable effort setting on a Haiku model. API pricing is US$0.10 per million input tokens and US$0.50 per million output tokens for prompts up to 100K tokens (US$0.50 and US$2.50 above that).ModelGemini 4 Argon30 SEP 2026 · GOOGLEGoogle announced Gemini 4 Argon and began rolling it out to trusted cyber defenders through its Fairwind Program. The model targets long-running coding, research and enterprise tasks with up to 1M output tokens; wider developer and consumer access is still planned.ModelOpenAI dots29 SEP 2026 · OPENAIOpenAI introduced dots, persistent agents with their own cloud computer and browser that can take on ongoing work between conversations. Powered by GPT-6 Astra, they can use connected apps, create files and delegate tasks to ChatGPT Work or Codex, with a gradual rollout to eligible accounts.Agent
§ 03 — SIDE BY SIDE

Every AI model release, seven labs.

OPENAIOpenAI: ChatGPT research preview — 2022-11-30OpenAI: GPT-4 reaches ChatGPT Plus — 2023-03-14OpenAI: GPT-4 Turbo and the Assistants API at DevDay — 2023-11-06OpenAI: GPT-4o — 2024-05-13OpenAI: GPT-4o mini replaces GPT-3.5 — 2024-07-18OpenAI: OpenAI o1-preview — 2024-09-12OpenAI: o3-mini, and reasoning for free users — 2025-01-31OpenAI: GPT-4.5 — 2025-02-27OpenAI: 4o image generation in ChatGPT — 2025-03-25OpenAI: GPT-4.1, 4.1 mini and 4.1 nano in the API — 2025-04-14OpenAI: o3 and o4-mini with full tool use — 2025-04-16OpenAI: gpt-oss open-weight models — 2025-08-05OpenAI: GPT-5 in ChatGPT — 2025-08-07OpenAI: GPT-5.1 in ChatGPT — 2025-11-12OpenAI: GPT-5.2 in ChatGPT — 2025-12-11OpenAI: ChatGPT Images — 2025-12-16OpenAI: GPT-5.3 Instant — 2026-03-03OpenAI: GPT-5.4 Thinking — 2026-03-05OpenAI: GPT-5.4 mini — 2026-03-18OpenAI: ChatGPT Images 2.0 — 2026-04-21OpenAI: GPT-5.5 in ChatGPT — 2026-04-23OpenAI: GPT-5.5 Instant — 2026-05-05OpenAI: GPT-5.6 Sol, Terra and Luna — 2026-07-09OpenAI: GPT-6 Astra — 2026-09-03OpenAI: ChatGPT Images 2.5 — 2026-09-08OpenAI: GPT-6 Sol and GPT-6 Luna — 2026-09-22OpenAI: GPT-6.1 Sol — 2026-09-29OpenAI: GPT-6.1 Terra — 2026-10-26OpenAI: Images 2.6 — 2026-11-04ANTHROPICAnthropic: Claude launch — 2023-03-14Anthropic: Claude 2 and claude.ai beta — 2023-07-11Anthropic: Claude 2.1 — 2023-11-21Anthropic: Claude 3 family — 2024-03-04Anthropic: Claude 3 Haiku — 2024-03-13Anthropic: Claude 3.5 Sonnet — 2024-06-21Anthropic: Computer use and upgraded Claude 3.5 models — 2024-10-22Anthropic: Claude 3.7 Sonnet and Claude Code — 2025-02-24Anthropic: Claude 4 — 2025-05-22Anthropic: Claude Opus 4.1 — 2025-08-05Anthropic: Claude Sonnet 4.5 — 2025-09-29Anthropic: Claude Haiku 4.5 — 2025-10-15Anthropic: Claude Opus 4.5 — 2025-11-24Anthropic: Claude Opus 4.6 — 2026-02-05Anthropic: Claude Sonnet 4.6 — 2026-02-17Anthropic: Project Glasswing and Claude Mythos Preview — 2026-04-07Anthropic: Claude Opus 4.7 — 2026-04-16Anthropic: Claude Opus 4.8 — 2026-05-28Anthropic: Claude Fable 5 and Mythos 5 — 2026-06-09Anthropic: Claude Sonnet 5 — 2026-06-30Anthropic: Claude Opus 5 — 2026-07-24Anthropic: Claude Fable 5.1 and Mythos 5.1 — 2026-09-01Anthropic: Claude Opus 5.5 — 2026-09-22Anthropic: Claude Sonnet 5.5 — 2026-09-28Anthropic: Claude Haiku 5.5 — 2026-10-07GOOGLEGoogle: Bard announced — 2023-02-06Google: PaLM 2 powers Bard — 2023-05-10Google: Gemini 1.0 — 2023-12-06Google: Gemini 1.5 Pro — 2024-02-15Google: Gemini 1.5 Flash — 2024-05-14Google: Gemini 2.0 Flash — 2024-12-11Google: Gemini 2.0 Flash Thinking Experimental — 2024-12-19Google: Gemini 2.0 Flash GA — 2025-02-05Google: Gemini 2.5 Pro Experimental — 2025-03-25Google: Gemini 2.5 Flash preview — 2025-04-17Google: 2.5 Deep Think — 2025-08-01Google: Nano Banana image editing — 2025-08-26Google: Gemini 3 Pro preview — 2025-11-18Google: Nano Banana Pro (Gemini 3 Pro Image) — 2025-11-20Google: Gemini 3 Flash — 2025-12-17Google: Gemini 3 Deep Think — 2026-02-12Google: Gemini 3.1 Pro — 2026-02-19Google: Gemini 3.5 Flash — 2026-05-19Google: Gemini Omni — 2026-05-19Google: Gemini 3.6 Flash and 3.5 Flash-Lite — 2026-07-21Google: Gemini 3.7 Flash — 2026-08-13Google: Gemini 3.8 Flash — 2026-09-02Google: Gemini 4 Argon — 2026-09-30Google: Gemini 4 Carbon — 2026-10-28Gemini 4 CarbonDEEPSEEKDeepSeek: DeepSeek LLM 7B and 67B — 2023-11-29DeepSeek: DeepSeek-V2 — 2024-05-06DeepSeek: DeepSeek-Coder-V2 — 2024-06-14DeepSeek: DeepSeek-V2.5 — 2024-09-05DeepSeek: DeepSeek-R1-Lite-Preview — 2024-11-20DeepSeek: DeepSeek-V3 — 2024-12-26DeepSeek: DeepSeek-R1 — 2025-01-20DeepSeek: Janus-Pro — 2025-01-27DeepSeek: DeepSeek-V3.1 — 2025-08-21DeepSeek: DeepSeek-V3.2-Exp — 2025-09-29DeepSeek: DeepSeek-V3.2 and V3.2-Speciale — 2025-12-01DeepSeek: DeepSeek-V4 Preview — 2026-04-24DeepSeek: DeepSeek-V4-Flash official release — 2026-07-31DeepSeek: DeepSeek-V4-Pro general availability — 2026-08-13DeepSeek: DeepSeek-V4.1-Flash — 2026-09-10ALIBABAAlibaba: Tongyi Qianwen unveiled — 2023-04-11Alibaba: Qwen-7B open weights — 2023-08-03Alibaba: Tongyi Qianwen 2.0 — 2023-10-31Alibaba: Qwen1.5 — 2024-02-04Alibaba: Qwen2 — 2024-06-07Alibaba: Qwen2.5 — 2024-09-19Alibaba: Qwen2.5-Coder 32B — 2024-11-12Alibaba: QwQ-32B-Preview — 2024-11-28Alibaba: Qwen2.5-VL — 2025-01-26Alibaba: Qwen2.5-Max — 2025-01-28Alibaba: QwQ-32B — 2025-03-06Alibaba: Qwen2.5-Omni — 2025-03-27Alibaba: Qwen3 — 2025-04-29Alibaba: Qwen3-Coder and Qwen Code — 2025-07-22Alibaba: Qwen3-Next — 2025-09-11Alibaba: Qwen3-Omni — 2025-09-22Alibaba: Qwen3-Max — 2025-09-24Alibaba: Qwen3-Max-Thinking — 2026-01-27Alibaba: Qwen3.5 and Qwen3.5-Plus — 2026-02-16Alibaba: Qwen3.6-Plus — 2026-04-02Alibaba: Qwen3.7-Max — 2026-05-20Alibaba: Qwen3.7-Plus — 2026-06-03Alibaba: Qwen3.8-Max — 2026-08-03Alibaba: Qwen3.8 open weights — 2026-08-12Alibaba: Qwen3.8-Flash-Next — 2026-08-26Z.AIZ.ai: ChatGLM and ChatGLM-6B — 2023-03-14Z.ai: ChatGLM2-6B — 2023-06-25Z.ai: ChatGLM3-6B — 2023-10-27Z.ai: GLM-4-9B open models — 2024-06-05Z.ai: GLM-4-0414, GLM-Z1 and Z.ai — 2025-04-14Z.ai: GLM-4.5 and GLM-4.5-Air — 2025-07-28Z.ai: GLM-4.5V — 2025-08-11Z.ai: GLM-4.6 — 2025-09-30Z.ai: GLM-4.6V — 2025-12-08Z.ai: GLM-4.7 — 2025-12-22Z.ai: GLM-5 — 2026-02-12Z.ai: GLM-5V-Turbo — 2026-04-02GLM-5V-TurboZ.ai: GLM-5.1 — 2026-04-07Z.ai: GLM-5.2 — 2026-06-16Z.ai: GLM-5.3 — 2026-08-14Z.ai: GLM-5.3-Flash — 2026-08-26MOONSHOT AIMoonshot AI: Kimi k1.5 — 2025-01-20Moonshot AI: Kimi K2 — 2025-07-11Moonshot AI: Kimi K2 Thinking — 2025-11-06Moonshot AI: Kimi K2.5 — 2026-01-27Moonshot AI: Kimi K2.6 — 2026-04-20Moonshot AI: Kimi K2.7 Code — 2026-06-12Kimi K2.7 CodeMoonshot AI: Kimi K3 — 2026-07-16Moonshot AI: Kimi K3.1 Pro — 2026-10-24
CURRENTSUPERSEDEDRETIREDLARGER SQUARE = WATERSHED
§ 04 — PACE

AI model releases per year.

Each announcement counts once, even when it covers several models (Claude 3’s Opus, Sonnet and Haiku are one release).

MODEL RELEASES BY YEAR
Major model releases per year by lab
YEAROPENAIANTHROPICGOOGLEDEEPSEEKALIBABAZ.AIMOONSHOT AI
2026131294865
202510685963
20243445510
20232331330
20221000000
§ 05 — QUESTIONS

LLM and AI model releases, answered.

OpenAI: Images 2.6 (4 Nov 2026), checked 6 Nov 2026. Google: Gemini 4 Carbon (28 Oct 2026), checked 6 Nov 2026. Moonshot AI: Kimi K3.1 Pro (24 Oct 2026), checked 6 Nov 2026. Anthropic: Claude Haiku 5.5 (7 Oct 2026), checked 6 Nov 2026. DeepSeek: DeepSeek-V4.1-Flash (10 Sep 2026), checked 6 Nov 2026. Alibaba: Qwen3.8-Flash-Next (26 Aug 2026), checked 6 Nov 2026. Z.ai: GLM-5.3-Flash (26 Aug 2026), checked 6 Nov 2026.

OpenAI, with 13 major model releases so far in 2026, against 12 from Anthropic, 9 from Google, 8 from Alibaba, 6 from Z.ai, 5 from Moonshot AI and 4 from DeepSeek. Each announcement counts once, even when it covers several models.

A new model, a new product or surface, a new plan tier, or the launch of a capability that changed how the product is used (memory, search, voice, agents). Incremental updates, settings, feature drops and regional rollouts are left out, and retirements are noted on the model they retired.

Each date is the day the maker announced the release on its own site or in its own documentation, and every entry links to that source. The timelines are checked against the official release notes regularly and whenever a new model reaches our benchmark tables.

Dates are the day the maker announced each release, taken from their own site and documentation.

Major releases only: new models, products, plans and capability launches. Retirements are noted on the model they retired.

Checked 6 Nov 2026 · How the current models compare →

MAKE SENSE OF THE NEXT RELEASE

Choose AI improvements around your business needs.