News
Releases, price changes, retirements, new capabilities and subscription plan changes, from each provider’s official pages and announcements. Summaries are ours; every item links to its source.
Nov 20, 2026
- xAI retires Grok Build 0.1 Grok Build 0.1 reached end of life on November 19, 2026, and is no longer served on the API.xAI
Nov 19, 2026
- OpenAI releases GPT Live 1.1 GPT Live 1.1 is available for existing Live session configurations and improves multilingual voice conversations. WebRTC sessions can use standby to pause inference during inactivity and resume when someone speaks.OpenAI
- Kling AI adds 4K and HDR output to Kling 4.0 APIs Kling 4.0 text-to-video, image-to-video, and Omni video APIs now accept 4K resolution. HDR output is available at 1080p and 4K through the dynamic range setting, with an additional fee per second of output video.Kuaishou Kling
- Google adds tool retrieval to the Interactions API The Interactions API can search a tool library and load function schemas on demand. Tool Retrieval works with Gemini 3.6 and later models, including all Gemini 4 models.Google
- Google releases Genie 4 Video model.Google
Nov 18, 2026
- Perplexity releases two contextual embedding models Perplexity added PPLX Embed V2 Context 9B and PPLX Embed V2 Context 0.6B to its Embeddings API, supporting text, chunked documents, and images. The 9b model costs $0.12 per million text tokens and $0.42 per million image tokens; the 0.6b model costs $0.02 and $0.40, respectively.Perplexity
- Google rolls out Genie 4 in Project Genie Google is rolling out Genie 4 through Project Genie to eligible Google AI Ultra subscribers. Google AI Pro members can join the waitlist; developers do not have access to the model.Google
- Google releases Gemini 4 Neon Gemini 4 Neon (gemini-4-neon) is generally available on the paid tier for $3 per million input tokens and $15 per million output tokens. It has a 1M-token context window and supports decoding up to 1M tokens.Google
- Google adds file processing and content slicing to Gemini API The Gemini API now accepts Microsoft Office and OpenDocument files and lets users select portions of documents, spreadsheets, videos, and audio in prompts and function responses. Supported selections include page ranges, worksheet tabs, cell ranges, time intervals, and content layers.Google
- GitHub adds Gemini 4 Neon to Copilot Gemini 4 Neon is generally available in GitHub Copilot for Pro+, Max, Business, and Enterprise users, with a gradual rollout. It is billed at provider list pricing under usage-based billing.GitHub
- Alibaba Qwen changes 8 models prices Qwen3.6 Plus (2026-04-02) on Alibaba Cloud Model Studio (USD): input $0.276 → $0.50 per 1M tokens. Qwen3.6 Flash (2026-04-16) on Alibaba Cloud Model Studio (USD): output $1.50 → $1.00 per 1M tokens. Qwen3.5 Flash (2026-02-23) on Alibaba Cloud Model Studio (USD): input $0.10 → $0.40, output $0.40 → $1.00 per 1M tokens.Alibaba Qwen
Nov 17, 2026
- OpenAI releases GPT-6.2 Astra and GPT-6.2 Sol GPT-6.2 Astra and GPT-6.2 Sol are available for demanding reasoning and coding tasks, and everyday coding and agent workflows, respectively. Both support the Responses API and Chat Completions; tool calling requires the Responses API.OpenAI
- OpenAI discounts background completions in the Responses API The Responses API public beta offers two-hour and eight-hour background completion windows, with model token rates 30% and 40% below Standard, respectively.OpenAI
- Kling AI releases Kling Image 4.0 API Kling Image 4.0 is available through text-to-image, image-to-image, and long-image APIs. Single images and image sets support up to native 8K resolution; the API also supports up to 10 combined image and subject references.Kuaishou Kling
- OpenAI cuts GPT-6 Rosalind prices by 33.3% GPT-6 Rosalind on OpenAI API (USD): input $15.00 → $10.00, output $75.00 → $50.00, cache read $1.50 → $1.00 per 1M tokens.OpenAI
- OpenAI releases GPT-6.2 Astra for Law On OpenAI API: $8.00 input, $40.00 output per 1M tokens.OpenAI
- Google adds voice editing to Gemini 3.5 Transcribe Gemini 3.5 Transcribe and Gemini 3.5 Transcribe Live now support voice-driven editing, rewriting, and formatting of dictated speech or existing text.Google
- GitHub adds MAI Code 2 models to Copilot MAI Code 2 and MAI Code 2 Flash are available to eligible annual-plan Copilot subscribers through usage-based billing. MAI Code 2 is offered on Pro, Pro+, Max, Business, and Enterprise; Flash is also available on Free and Student.GitHub
Nov 16, 2026
- OpenAI releases CSAM Image Classifier for approved organizations Approved organizations can use the OpenAI CSAM Image Classifier at no cost to prioritize images for human review. It returns separate signals for potential CSAM and potential AI-generated CSAM.OpenAI
- OpenAI makes healthcare computer-use agents available Healthcare teams, developers, and clinicians can use computer-use agents through the OpenAI API, ChatGPT for Healthcare, and ChatGPT for Clinicians. OpenAI reports GPT-6 Astra scored 94.3% on HealthAdminBench and GPT-6 Luna scored 91.8%.OpenAI
- OpenAI expands computer-use capabilities for healthcare OpenAI says computer-use agents and improved clinical search are available today in ChatGPT Work and Codex, subject to workspace settings, and computer-use agents are also available through the API. The announcement describes more capable, safe, and affordable models but names no new model or API limits.OpenAI
- GitHub retires three models from Copilot GitHub deprecated Claude Opus 4.8, Claude Opus 5 and GPT-5.6 Terra across most Copilot experiences on November 16, 2026. Claude Opus 4.8 remains available to individual subscribers on annual plans, and Claude Opus 5 remains available for runtime refusal fallback through existing authorized access.GitHub
- Qwen releases Qwen4 Max Qwen introduces Qwen4 Max, built on the Qwen4 architecture with native vision capabilities and a one-million-token context window.Alibaba Qwen
Nov 15, 2026
- ByteDance releases Dreamina Seedance 2.5 Flash Dreamina Seedance 2.5 Flash supports text to video, frame guided video generation, omni reference to video, editing, and extension at 480p and 720p resolutions. It does not support Draft mode.ByteDance Seed
- ByteDance releases Dreamina Seedance 2.5 Dreamina Seedance 2.5 can generate videos up to 30 seconds per request, extend videos over multiple rounds, use up to 50 omni reference assets per request, and edit at specified timestamps.ByteDance Seed
Nov 13, 2026
- xAI expands Grok Imagine Image 2 editing and aspect ratios Grok Imagine Image 2 now accepts up to five source images per editing request, up from three, and supports 21:9 and 5:2 aspect ratios. Its default quality setting is now auto, which uses low for generation and medium for editing; images are billed at the quality served.xAI
- DeepSeek adds application tools to Flash APIs DeepSeek adds web search, multi-step web search, file input, URL reading and safety checks to its Messages, Responses and Chat Completions APIs for deepseek-flash. Search and Fetch, OCR, ASR and TTS are also available as standalone APIs.DeepSeek
Nov 12, 2026
- Perplexity releases Decider v1.3 models Perplexity releases PPLX Decider V1.3 (27B) and PPLX Decider Flash V1.3 (4B), with input prices of $0.02 and $0.01 per million tokens, respectively; output tokens remain free. Both models support text, images and grouped questions, while the new locate question type is available only on v1.3 models and does not work with web grounding.Perplexity
- Moonshot AI releases Kimi K2.7 Code models Moonshot AI released Kimi K2.7 Code and later its HighSpeed variant through the Open Platform API. The platform also added custom top-up amounts.Moonshot AI
- Moonshot AI releases Kimi K2.6 Moonshot AI released Kimi K2.6 through its Open Platform API.Moonshot AI
- Google releases Gemini 4 Hydrogen Gemini 4 Hydrogen (gemini-4-hydrogen) is generally available, including on the free tier; paid pricing is $1.50 per million input tokens and $7.50 per million output tokens. It adds an xhigh thinking level.Google
- Google makes Gemini 4 Hydrogen available to developers Developers can build with Gemini 4 Hydrogen through the Gemini API in Google AI Studio and Android Studio. The model offers four configurable thinking levels and supports coding, debugging, web development, and multimodal workflows.Google
- Cognition updates Devin’s Teams plan Teams users get two concurrent free SWE-2 sessions per user, and free SWE-2 becomes a permanent part of Teams, Pro, and Max. The plan adds Devin Desktop access for everyone, SSO and cost controls; seat limits fall from 200 to 100, and quota seats are retired in 30 days.Cognition
Nov 11, 2026
- Runway releases Praxis-1 with open weights Runway has made Praxis-1 available to everyone under the Apache 2.0 license, including its weights and inference and fine-tuning code. Developers can access it on GitHub, Hugging Face, vLLM-Omni and Runway Dev.Runway
- Alibaba Qwen releases qwen4-max-preview Language model.Alibaba Qwen
- Google releases Gemini Decision and Decisions API Google released Gemini Decision (gemini-decision-1) and the Decisions API for binary, categorical, and rubric decisions across text, images, audio, video, and PDFs, with zero output tokens.Google
- Baidu releases ERNIE Light DT 1.0 1M-token context window.Baidu
- ElevenLabs cuts 3 models prices Scribe V2 Medical on ElevenLabs API (USD): audio_second $0.22 → $0.20 per 3600. Scribe V2 Realtime on ElevenLabs API (USD): audio_second $0.39 → $0.22 per 3600.ElevenLabs
- Baidu releases ERNIE-5.5 Flash Preview Baidu made ERNIE-5.5 Flash Preview available through the Qianfan API on November 11, 2026. The model has 280B total parameters, 14B active parameters and a 1M-token context; its weights are planned for Apache 2.0 release soon.Baidu
Nov 10, 2026
- OpenAI deprecates GPT Image 2 GPT Image 2 is now deprecated and shuts down on 2027-02-10.OpenAI
- MiniMax releases MiniMax-H3 Image Highspeed Image model.MiniMax
- MiniMax releases Image-01 Live Image model.MiniMax
Nov 9, 2026
- Meta releases Muse Ember Language model: text, image, audio, video and PDF in, text out.Meta
- Google introduces Gemini Robotics 2.5 Google introduced Gemini Robotics 2.5, a vision-language-action model rolling out to its robotics partners. It was evaluated across more than 200 tasks, but the announcement does not say developers can access it.Google
- Zhipu AI GLM Coding Plan: no longer lists the Team Standard Seat tier and 1 more change No longer lists the Team Standard Seat tier; no longer lists the Team Premium Seat tier.Zhipu AI
- ElevenLabs releases Scribe V3 and Scribe V3 Live Scribe V3 and Scribe V3 Live are available in the Speech to Text API for recorded and real-time transcription, respectively. They add transcription features, and Scribe V3 reports 27% fewer word errors than Scribe v2 while Scribe V3 Live reports a 44% lower semantic word error rate than Scribe v2 Realtime on FLEURS.ElevenLabs
- ByteDance Seed introduces Seedream 5.5 ByteDance Seed introduces Seedream 5.5, an image creation model with native transparent RGBA generation and more stable multi-turn editing. The announcement says a Pro version will follow.ByteDance Seed
- Anthropic launches Claude Fable 5.2 and Claude Mythos 5.2 Anthropic launched Claude Fable 5.2, available on the Claude API, Claude Platform on AWS, and Claude Models on Google Cloud. Claude Mythos 5.2 offers the same capabilities only to organizations verified through Anthropic’s verification programs; both models have a 1 million token context window, 128,000 max output tokens, and cost $8 per million input tokens and $40 per million output tokens.Anthropic
- Anthropic adds fast mode for Claude Fable 5.2 Fast mode for Claude Fable 5.2 is available in Messages API requests, delivering up to 2.5 times higher output tokens per second. It costs $16 per million input tokens and $80 per million output tokens.Anthropic
Nov 6, 2026
- Google deprecates Gemini Embedding 2 Preview Effective November 6, 2026, requests to gemini-embedding-2-preview route to gemini-embedding-2 at its rates. The replacement uses the same API schema, pricing, and embedding space.Google
Nov 5, 2026
- OpenAI makes GPT-6-Cyber available OpenAI says developers can use GPT-6-Cyber to investigate vulnerabilities, reproduce them in an authorized test environment, and verify fixes. The entry directs users to Daybreak setup before their first request.OpenAI
- Amazon Web Services Kiro: adds the Credit pooling for teams of 2 or more tier at $20.00/month Adds the Credit pooling for teams of 2 or more tier at $20.00/month.Amazon Web Services
- Google adds multimodal function responses to its APIs GenerateContent and Interactions APIs can now pass audio, video, and Office and OpenDocument files in function responses. Responses can also include Files API URIs, with video processing options in some API versions.Google
- Google releases Gemini Omni 1.2 Flash Language model.Google
- Ant Group releases Ling-3.2 Flash Ant Group launched Ling-3.2 Flash with a 1M-token context window and image, text, and video input. It is available through the API using OpenAI or Anthropic protocols and is free for the first two weeks after launch.Ant Group
Nov 4, 2026
- StepFun raises StepAudio 3 ASR Max prices by 50% StepAudio 3 ASR Max on StepFun API (USD): audio_second $0.40 → $0.60 per 3600.StepFun
- StepFun raises StepAudio 3 ASR Max prices by 167.9% in China StepAudio 3 ASR Max on StepFun API (CNY): audio_second ¥2.80 → ¥7.50 per 3600.StepFun
- Perplexity adds web grounding to Decisions API The Decisions API can run a Fast Search for each question, up to 100 questions, and include the results as evidence before the model answers. Each search costs $0.001 in addition to input tokens for the added results.Perplexity
- OpenAI releases GPT Image 2.6 Sunburst and Flare OpenAI released GPT Image 2.6 Sunburst and GPT Image 2.6 Flare for image generation and editing, with adjustable complexity, SVG and animated GIF generation. Markup-guided editing was also added for supported GPT Image 2.5 and 2.6 models.OpenAI
- OpenAI brings GPT-5.6 Cyber to AWS GovCloud GPT-5.6 Cyber is now available to approved customers through Amazon Bedrock in AWS GovCloud (US-East) for authorized cybersecurity work.OpenAI
- Kling AI releases Kling 4.0 Flash API Kling 4.0 Flash API is available for text-to-video, image-to-video and Omni reference-to-video, generating videos from 3 to 20 seconds at 720P. Prompts support up to 26,000 English characters, 10 reference images and 5 reference videos.Kuaishou Kling
- Google adds voice remixing to Gemini 3.8 TTS Developers can use natural-language instructions to create an independent stored voice from a prebuilt or designed voice, then synthesize speech with Gemini 3.8 Flash TTS or Gemini 3.8 Flash-Lite TTS.Google
Nov 3, 2026
- ByteDance Seed TRAE: no longer lists the Lite tier and 2 more changes No longer lists the Lite tier; no longer lists the Pro+ tier; no longer lists the Ultra tier.ByteDance Seed
Nov 2, 2026
- xAI retires Grok Imagine Image Quality Requests to grok-imagine-image-quality and grok-imagine-image-pro are now served by grok-imagine-image-2.0-low, using grok-imagine-image-2.0 with quality set to low at a lower per-image price.xAI
- StepFun raises 3 models prices StepAudio 2.5 ASR on StepFun API (USD): audio_second $0.022 → $0.05 per 3600. StepAudio 2.5 ASR Stream on StepFun API (USD): audio_second $0.18 → $0.40 per 3600. Step ASR 1.1 Stream on StepFun API (USD): audio_second $0.37 → $0.81 per 3600.StepFun
- OpenAI makes outbound SIP calling available through the Live API Organizations can now connect GPT-Live to a phone number through their SIP provider using the Live API.OpenAI
- Google releases Gemini 4 Argon and announces Gemini 4 Carbon Google rolled out Gemini 4 Argon globally across Google AI Studio, the Gemini API, Google Cloud, Search and the Gemini app. Gemini 4 Carbon is in pre-release testing with the U.S. government and selected cyber defenders, with a broader launch planned.Google
- Google deprecates Gemini 2.5 Pro Gemini 2.5 Pro is deprecated and can no longer be selected in Google AI Studio. Starting December 1, 2026, API requests will route to Gemini 3.8 Flash and use its rates.Google
- Google deprecates Gemini 2.5 Flash Gemini 2.5 Flash is deprecated and can no longer be selected in Google AI Studio. Starting December 1, 2026, API requests will route to Gemini 3.5 Flash-Lite and use its rates.Google
- ElevenLabs releases SFX v3 for sound effects ElevenLabs made SFX v3 (eleven_text_to_sound_v3) available to every workspace as its newest sound effects model. The company says it follows multi-instruction prompts more closely and handles music and vocals better; eleven_text_to_sound_v2 will be removed on March 1, 2027.ElevenLabs
- Amazon makes OpenAI GPT-6.1 Terra generally available on Bedrock OpenAI GPT-6.1 Terra is generally available on Amazon Bedrock. It accepts up to 1 million input tokens with text and image input.Amazon Web Services
Nov 1, 2026
- Meituan makes LongCat-2.5 available Meituan says LongCat-2.5 is now available. It supports multimodal visual coding, creative design, 3D tasks, and complex software engineering across several agent harnesses.Meituan
Oct 31, 2026
- StepFun raises 5 models prices in China StepAudio 2.5 Realtime on StepFun API (CNY): input ¥10.00 → ¥13.00, output ¥70.00 → ¥90.00, cache read ¥2.00 → ¥3.00 per 1M tokens. StepAudio 2.5 Chat on StepFun API (CNY): input ¥10.00 → ¥13.00, output ¥25.00 → ¥32.00, cache read ¥2.00 → ¥3.00 per 1M tokens. StepAudio 2.5 ASR on StepFun API (CNY): audio_second ¥0.15 → ¥0.45 per 3600. StepAudio 2.5 ASR Stream on StepFun API (CNY): audio_second ¥1.20 → ¥4.20 per 3600. Step ASR 1.1 Stream on StepFun API (CNY): audio_second ¥2.60 → ¥5.50 per 3600.StepFun
- StepFun releases StepAudio 3 Realtime Realtime voice model.StepFun
- StepFun releases StepAudio 3 Music Music model.StepFun
- StepFun releases StepAudio 3 Gen Text-to-speech model.StepFun
- Moonshot AI adds Kimi K3.1 Pro and K3 to Batch API Batch API now supports kimi-k3.1-pro and kimi-k3 on both the CN and Global platforms. Batch inference costs 60% of standard rates; these models also incur Cache Write fees, and cache hits are best-effort.Moonshot AI
- MiniMax M Plan: Go 11 → 22 USD / month and 5 more changes Go 11 → 22 USD / month; Go 24.5 → 49 CNY / month (china); Explore 27.5 → 55 USD / month; Explore 59.5 → 119 CNY / month (china); Build 66 → 132 USD / month; Build 234.5 → 469 CNY / month (china).MiniMax
- OpenAI releases GPT-6 Astra for Law On OpenAI API: $10.00 input, $50.00 output per 1M tokens.OpenAI
Oct 30, 2026
- OpenAI releases GPT-6 Rosalind for approved research organizations OpenAI released GPT-6 Rosalind for biological reasoning and scientific tool use. Access is limited to organizations approved through the Rosalind Research trusted-access program.OpenAI
- Mistral AI cuts Mistral Large 4 prices Mistral Large 4 on Mistral AI API (USD): input $1.36 → $0.70, output $4.18 → $2.10, cache read $0.14 → $0.07 per 1M tokens.Mistral AI
- Google makes Computer Use available across two APIs Google made the Computer Use tool generally available in the Interactions and GenerateContent APIs.Google
- Anthropic Claude plans: adds the Ultra tier at $200.00/month and 1 more change Adds the Ultra tier at $200.00/month; no longer lists the Max 20x tier.Anthropic
- Anthropic retires Claude Opus 4.8 fast mode Anthropic will remove fast mode for Claude Opus 4.8 on November 30, 2026. Requests using fast mode after removal will return an error; Anthropic says to migrate to Claude Opus 5.5 to continue using it.Anthropic
Oct 29, 2026
- Tencent previews upcoming Hy4 and Hy4 Flash releases Tencent says Hy4’s general release is coming soon, with improved performance on long-horizon productivity tasks, reasoning efficiency and visual understanding. Hy4 Flash is also in preparation; the announcement does not say either model is available to developers yet.Tencent
Oct 28, 2026
- Kuaishou Kling releases Multi Element Video Editing Video model.Kuaishou Kling
- Kling AI expands Kling 4.0 video API capabilities Kling 4.0 API supports single-run video generation up to 30 seconds, 720p and 1080p, plus 21:9 and adaptive aspect ratios. It adds up to 10 keyframes, voice references for Omni video, and higher asset limits, including 26,000-character English prompts.Kuaishou Kling
- Google announces Gemini 4 Carbon Google announced Gemini 4 Carbon, which will initially be available to a subset of cyber defenders before wider access. Its standard prices after the introductory period are $4 per million input tokens and $20 per million output tokens.Google
- GitHub makes Copilot Fusion generally available Copilot Fusion is generally available in the model picker, where it routes and orchestrates models for each task. Users can choose Standard, High, or Max effort levels.GitHub
- Black Forest Labs adds custom output sizes to FLUX 3 Image The FLUX 3 Image API now accepts custom width and height up to 16.8MP, or `resolution: "match_input"` for edits up to 4MP. Single-reference-image edits also follow instructions more closely; pricing is unchanged.Black Forest Labs
- Kiro expands its student plan to 41 universities Kiro for Students is expanding to 41 more universities. Eligible students get one free year with 1,000 monthly credits and access to models available in paid plans.Amazon Web Services
Oct 27, 2026
- Alibaba Qwen releases Qwen Audio 3.1 Realtime Flash Realtime voice model.Alibaba Qwen
- Google makes Gemini Robotics ER 2 available in preview Gemini Robotics ER 2 is available in public preview on the Gemini platform, including through Model Garden. Developers and enterprise teams can start using it now.Google
- Google announces Genie 4 world model preview Google announced Genie 4, a real-time interactive world model that generates native audio and can be steered with natural language. It is available only to trusted testers as a research preview; access through Project Genie and the Gemini API and Google AI Studio is coming soon.Google
- Ant Group releases Ming image models Ant Group launched Ming Image 0.1 Design and Ming Image 0.1 Design Layer, available through its open API.Ant Group
- Amazon offers Canada inference for Claude Haiku 5.5 Amazon Bedrock now supports Canada cross-Region inference for Anthropic’s Claude Haiku 5.5, routing requests between ca-central-1 and ca-west-1. This lets customers keep inference processing within Canada.Amazon Web Services
- Kiro changes its free trial and prorates plans Starting November 1, first-time upgrades from the free tier get the rest of the month free and 500 trial credits, which expire after 12 months. Plan fees and credits are prorated for partial months; the previous $20 discount remains for upgrades before November 1.Amazon Web Services
Oct 26, 2026
- OpenAI releases GPT-6.1 Terra OpenAI released GPT-6.1 Terra for the Responses API and Chat Completions API.OpenAI
- OpenAI makes GPT-6.1 Terra available in Work and Codex GPT-6.1 Terra is available in ChatGPT Work and Codex for Plus, Pro, Business, Enterprise, and Edu plans. It is not available in Chat, and Enterprise access depends on workspace availability and administrator controls.OpenAI
- Kiro adds Claude Haiku 5.5 to its IDE, CLI and web Claude Haiku 5.5 is available with experimental support to Kiro Pro, Pro+, Pro Max and Power customers in us-east-1 and eu-central-1. It has a 1M-token context window and a 0.06x credit multiplier; requests exceeding 100K tokens use a 0.3x multiplier.Amazon Web Services
Oct 24, 2026
- MiniMax releases MiniMax-H3 Max Turbo MiniMax-H3 Max Turbo is now available.MiniMax
- MiniMax releases MiniMax-H3 Lite MiniMax-H3 Lite is now available.MiniMax
- MiniMax releases MiniMax-H3 Highspeed MiniMax-H3 Highspeed is now available.MiniMax
Oct 23, 2026
- OpenAI retires o4 Mini, o3 Mini, o1 and 12 more o4 Mini, o3 Mini, o1, o1 Pro, GPT Image 1, GPT-4o (2024-05-13), GPT-4, GPT-4 Turbo, GPT-4 1106 Preview, GPT-4.1 Nano, GPT-3.5 Turbo, GPT-4 Turbo (2024-04-09), GPT-4.1 Nano (2025-04-14), GPT-4 (2023-06-13) and GPT-3.5 Turbo (2024-01-25) are now retired.OpenAI
- OpenAI adds scheduled automations to the Agents API The Agents API public beta now supports one-time and recurring scheduled tasks in new or existing sessions. Agents can also schedule follow-ups when the automations tool is enabled.OpenAI
- Moonshot AI releases Kimi K3.1 Pro models Kimi K3.1 Pro and Kimi K3.1 Pro Highspeed are available today through the Kimi API and Kimi products. K3.1 Pro has a 1-million-token context window; full weights are scheduled for release by Nov. 4, 2026.Moonshot AI
- Amazon adds domain and location controls to Bedrock Web Search AWS Bedrock Web Search now supports domain filtering and optional user location in OpenAI Responses API requests. The controls are available in US East (N. Virginia), US East (Ohio), US West (Oregon), and AWS GovCloud (US-West).Amazon Web Services
Oct 22, 2026
- OpenAI makes Decisions API generally available for GPT-6 Luna The Decisions API is generally available with gpt-6-luna. Requests can include up to 400 questions, up from 200 in public beta, and support shared-input caching, images in question instructions, and structured data through input_json.OpenAI
- Google retires Veo 3.1, Veo 3.1 Lite, Veo 3.1 Fast and 1 more Veo 3.1, Veo 3.1 Lite, Veo 3.1 Fast and Gemini Omni Flash Preview are now retired.Google
- Zhipu AI releases GLM-5.3 FlashX Legal On Z.ai API: $0.37 input, $1.25 output per 1M tokens.Zhipu AI
- Anthropic launches Claude Models on Google Cloud Anthropic-managed Claude API access is now available through Google Cloud projects. Google Cloud IAM manages access, and Google Cloud Marketplace handles billing.Anthropic
- Anthropic adds Opus 5.5 fast mode on Google Cloud Fast mode for Claude Opus 5.5 is now available through Anthropic-managed Claude Models on Google Cloud.Anthropic
Oct 21, 2026
- Xiaomi retires MiMo-V2.5 and MiMo-V2.5 Pro MiMo-V2.5 and MiMo-V2.5 Pro are now retired.Xiaomi
- Runway adds MOV output for Seedance 2.5 Seedance 2.5 can now return MOV files through the text to video, image to video, and video to video endpoints. MOV costs the same as MP4; draft previews and enhanced drafts remain MP4.Runway
- Tencent releases Hy Nomas Video model.Tencent
- Cohere releases Rerank 5 Pro and Fast Cohere has released rerank-v5.0-pro and rerank-v5.0-fast, with a 32k-token context window and support for over 100 languages. An optional Context Focus capability returns relevant passages as scored character spans in the same rerank call.Cohere
Oct 20, 2026
- Mistral AI raises Voxtral TTS prices Voxtral TTS on Mistral AI API (USD): character $0.00 → $16.00 per 1000000.Mistral AI
- Mistral AI releases Voxtral TTS 3 Voxtral TTS 3 is now available.Mistral AI
- Mistral AI releases Voxtral Small Transcribe 3 Voxtral Small Transcribe 3 is now available.Mistral AI
- Mistral AI releases Voxtral Mini Transcribe Realtime 3 Voxtral Mini Transcribe Realtime 3 is now available.Mistral AI
- Mistral AI releases Voxtral Mini Transcribe 3 Voxtral Mini Transcribe 3 is now available.Mistral AI
- OpenAI adds 12-hour prompt caching to the API The API now offers an option to cache prompts for 12 hours, extending the time shared instructions, tools, and context remain cached. Writes to the 12-hour cache cost 2.5 times the input-token price.OpenAI
- Mistral AI releases Voxtral 3 speech models Mistral makes Voxtral 3 speech models available through its API, including transcription models and Voxtral Speech 3 for speech generation. Listed prices range from $0.0035 to $0.008 per minute for transcription and $13 per 1M characters for speech generation.Mistral AI
- MiniMax M Plan: Go 22 → 11 USD / month and 5 more changes Go 22 → 11 USD / month; Go 49 → 24.5 CNY / month (china); Explore 55 → 27.5 USD / month; Explore 119 → 59.5 CNY / month (china); Build 132 → 66 USD / month; Build 469 → 234.5 CNY / month (china).MiniMax
Oct 19, 2026
- Perplexity raises Agent API xhigh output limit The Agent API’s xhigh preset now supports up to 512,000 output tokens, up from 128,000.Perplexity
- MiniMax makes M3.1-Flash available MiniMax-M3.1 Flash is now available through the API. It supports a context window of up to 1,000,000 tokens and accepts text, image, and video inputs.MiniMax
- Ant Group releases Ming-3.0 TTS 0.5B Text-to-speech model.Ant Group
- Inception releases Mercury 3 Mercury 3 is available on Inception’s API Platform, OpenRouter, Baseten, and Vercel. It has a 1M-token context and standard pricing of $0.20 per million input tokens and $0.70 per million output tokens; launch pricing is 50% off.Inception
- GitHub deprecates six Copilot models On October 19, 2026, GitHub deprecated Gemini 3.7 Flash, GPT-5.5, GPT-5.4, GPT-5.4 mini, GPT-5 mini, and Grok 4.5 across most Copilot experiences. Suggested alternatives are Gemini 3.8 Flash, GPT-6.1 Sol, GPT-6 Luna, and Grok 4.7.GitHub
- ElevenLabs releases Eleven Realtime V1 Mini and Max Eleven Realtime V1 Mini and Max are new voice agent models rolling out gradually across workspaces. They combine LLM, speech-to-text and text-to-speech settings in one configuration.ElevenLabs
- Ant Group releases three Ming audio models Ant Group launched Ming-3.0 ASR Tiny, Ming-3.0 TTS and Ming-3.0 Audio Flash, available through its open API.Ant Group
Oct 17, 2026
- Cohere releases North Small Translate 09-2026 North Small Translate 09-2026 is now available.Cohere
- DeepSeek upgrades V4.1 Flash to version 1017 DeepSeek upgraded DeepSeek V4.1 Flash to DeepSeek-V4.1-Flash-1017; developers can access it using deepseek-flash, with the API method and pricing unchanged.DeepSeek
Oct 16, 2026
- OpenAI adds prompt-generated voices to GPT-Live OpenAI now lets all API customers create reusable GPT-Live voices from text descriptions through the API or voice Playground. Voice creation is free; speech generation uses regular GPT-Live session billing, and nine voices in German, French, Hindi, Japanese, and Korean were added.OpenAI
- Ant Group cuts Ling-3.1 Flash cache read prices by 20% Ling-3.1 Flash on Ant Ling API (CNY): cache read ¥0.125 → ¥0.10 per 1M tokens.Ant Group
Oct 15, 2026
- OpenAI integrates Decisions with GPT-Live The v1/live/sessions API now supports native delegation from GPT-Live to Decisions, as well as triggering decisions as input or output transcript deltas arrive. The integration covers gpt-live-1 and gpt-6-luna.OpenAI
- Mistral AI releases Timestral forecasting models Mistral AI released Timestral, a family of time-series models with 5M, 28M, 123M and 209M parameters. The models support probabilistic forecasting, and their context length is 32,768 time steps.Mistral AI
- MiniMax Token Plan: Plus is closed to new subscriptions and 2 more changes Plus is closed to new subscriptions; Max is closed to new subscriptions; Ultra is closed to new subscriptions.MiniMax
- Meta releases Muse Spark Open weights Meta released the open weights for Muse Spark Open, an 816B-parameter multimodal model with a 1M-token context window. It is available today through Hugging Face and API platforms including Baseten, Fireworks AI, Groq, Modal, Nebius, OpenRouter and Together AI.Meta
- GitHub deprecates selected Copilot models on November 16 GitHub will deprecate Claude Opus 4.8, Claude Opus 5, and GPT-5.6 Terra across Copilot experiences on November 16, 2026. Claude Opus 4.8 remains available to individual Copilot subscribers on annual plans; suggested replacements are Claude Opus 5.5, GPT-6.1 Sol, or GPT-6 Luna.GitHub
- Anysphere Cursor: adds the Free tier (free) and 4 more changes Adds the Free tier (free); adds the Plus tier at $8.00/month; adds the Premium tier at $30.00/month; adds the Super tier at $100.00/month; no longer lists the Teams Premium tier.Anysphere
- Black Forest Labs releases FLUX 3 Image Klein Black Forest Labs released FLUX 3 Image Klein, a 5.6B-parameter open-weights model for image generation and editing, available through its API and as downloadable weights. API pricing starts at $0.009 per 1K image.Black Forest Labs
- Anthropic makes Opus 5.5 fast mode available on Azure Fast mode for Claude Opus 5.5 is now available in Claude on Microsoft Foundry for deployments hosted on Azure.Anthropic
- Ant Group raises Ling-3.1 Flash and Ling-3.0 Flash prices Ling-3.1 Flash on Ant Ling API (CNY): input ¥0.00 → ¥0.50, output ¥0.00 → ¥1.50, cache read ¥0.00 → ¥0.125 per 1M tokens. Ling-3.0 Flash on Ant Ling API (CNY): input ¥0.125 → ¥0.14, output ¥0.375 → ¥0.42, cache read ¥0.025 → ¥0.028 per 1M tokens.Ant Group
Oct 14, 2026
- xAI releases Grok Voice 3.0 Grok Voice 3.0 is available for Speech to Speech at the same price as Grok Voice 2.0. The grok-voice-latest alias now routes to Grok Voice 3.0, upgrading sessions that use it automatically.xAI
- xAI deprecates Grok Voice 2.0 xAI deprecated grok-voice-2.0 on October 14, 2026. Developers can migrate to grok-voice-3.0 or grok-voice-latest at the same price.xAI
- TypeSafe releases Jev 1.14.0 preview TypeSafe released Jev 1.14.0 through jev-preview, with small improvements and an expected larger gain in multilingual support. The release also enables experimental image support and longer context through API flags; the new box type is available only on Jev 1.14.0.TypeSafe
- IBM releases Granite Switch 4.1 8B Granite Switch 4.1 8B is now available.IBM
- GitHub introduces HydraFusion Max for Copilot GitHub adds HydraFusion Max alongside HydraFusion in research preview, available across VS Code, the GitHub Copilot app, and Copilot CLI. Both are offered on Copilot Pro, Pro+, Business, and Enterprise plans when experimental or preview features are enabled, subject to organization model policies.GitHub
Oct 12, 2026
- xAI launches The X Subscription The X Subscription combines Grok, Cursor, and X, with new customers able to sign up starting October 12, 2026. Plans range from free to $200 per month; existing SuperGrok, Cursor, and X Premium plans remain available to current subscribers but are closed to new sign-ups.xAI
- StepFun raises Step 5 Preview cache read prices by 100% in China Step 5 Preview on StepFun API (CNY): cache read ¥0.35 → ¥0.70 per 1M tokens.StepFun
- Meta releases Muse Wonder models Muse Wonder is available through the Meta Model API and Muse Code. Muse Wonder Full is available to Muse Code subscribers.Meta
- ElevenLabs adds audio effects to speech generation Text to Speech and Text to Dialogue requests can now apply audio effects, including reverb and stereo panning. The realtime Speech to Text WebSocket also adds turn prediction as a commit strategy.ElevenLabs
Oct 11, 2026
- Alibaba Qwen cuts Fun ASR Flash 8K Realtime (2026-01-28) prices by 33.3% Fun ASR Flash 8K Realtime (2026-01-28) on Alibaba Cloud Model Studio (CNY): audio_second ¥0.00033 → ¥0.00022.Alibaba Qwen
Oct 10, 2026
- StepFun raises Step 5 Preview cache read prices by 100% Step 5 Preview on StepFun API (USD): cache read $0.05 → $0.10 per 1M tokens.StepFun
- Alibaba Qwen changes Kimi K2.7 Code prices Kimi K2.7 Code on Alibaba Cloud Model Studio (USD): input $0.894 → $0.8939, output $3.713 → $3.7131 per 1M tokens.Alibaba Qwen
Oct 9, 2026
- Alibaba Qwen releases Qwen Image 2.1 Turbo Qwen Image 2.1 Turbo is now available.Alibaba Qwen
- Perplexity replaces its Decisions API model and cuts price Perplexity replaced PPLX Decider V1 27B with PPLX Decider V1.1 27B for the Decisions API. Input pricing fell from $0.04 to $0.02 per million tokens; requests using the older ID continue to work at the new rate.Perplexity
Oct 8, 2026
- OpenAI adds Ultrafast mode for GPT-6.1 Sol GPT-6.1 Sol can use Ultrafast mode in the Responses API to reduce the time between generated output tokens. It is available to all API users, subject to rate limits, with global processing and US and EU data residency.OpenAI
- Mistral AI Mistral plans: Pro usage limits change (month usd) Pro usage limits change (month usd).Mistral AI
- Google retires Deep Research Agent preview Google will shut down deep-research-pro-preview-12-2025 on October 23, 2026. It recommends migrating to deep-research-preview-04-2026 or deep-research-max-preview-04-2026.Google
- Google deprecates Gemini 3.7 Flash Google deprecated gemini-3.7-flash and replaced it with gemini-3.8-flash. Requests to the deprecated model are automatically routed to the replacement.Google
- Google deprecates Gemini 3.5 Flash Google deprecated gemini-3.5-flash and replaced it with gemini-3.6-flash. Requests to the deprecated model are automatically routed to the replacement.Google
- OpenAI ChatGPT plans for Codex: adds the Pro $500 tier at $500.00/month Adds the Pro $500 tier at $500.00/month.OpenAI
- Anthropic cuts Claude Sonnet 5.5 prices by 50% Claude Sonnet 5.5 on Claude API (USD): cache read $0.20 → $0.10 per 1M tokens.Anthropic
- Amazon Web Services cuts Claude Sonnet 5.5 cache read prices by 50% Claude Sonnet 5.5 on Amazon Bedrock (USD): cache read $0.22 → $0.11 per 1M tokens. Claude Sonnet 5.5 on Amazon Bedrock (USD): cache read $0.20 → $0.10 per 1M tokens.Amazon Web Services
Oct 7, 2026
- Perplexity adds Claude Haiku 5.5 to Agent API The Agent API now supports Anthropic’s claude-haiku-5-5. Requests with more than 100,000 input tokens are charged at higher rates.Perplexity
- Anthropic releases Claude Haiku 5.5 Claude Haiku 5.5 is available now across platforms, including AWS, Google Cloud, and Microsoft Azure, with API access under claude-haiku-5-5. Pricing per million tokens for prompts up to 100,000 tokens is $0.10 input and $0.50 output; above that, it is $0.50 input and $2.50 output.Anthropic
- Anthropic cuts Claude Sonnet 5.5 cache-read prices Anthropic lowered prompt cache-read pricing for Claude Sonnet 5.5 from $0.20 to $0.10 per million tokens, or 0.05 times the base input price instead of 0.1 times. Cache writes and other prices are unchanged.Anthropic
Oct 6, 2026
- OpenAI launches Decisions API beta for GPT-6 Luna OpenAI released the Decisions API in beta with GPT-6 Luna. It accepts text and images and returns typed answers; the announcement says it is 10 times faster than the Responses API.OpenAI
- Mistral AI previews Mistral Large 4 Mistral Large 4 is available in public preview through the Mistral Studio API. Mistral says it will release the model weights by the end of October 2026; listed API pricing is $1.36 per million input tokens and $4.18 per million output tokens.Mistral AI
- Google schedules Gemini 3.1 Flash Image shutdown Google deprecated gemini-3.1-flash-image, with shutdown scheduled for October 29, 2026. The announcement recommends migrating to gemini-nano-banana-2.1.Google
- Google releases EmbeddingGemma 2 Google launched EmbeddingGemma 2, a 740-million-parameter multimodal embedding model for text, code, images, video and audio, under the Apache 2.0 license. It has an 8K-token context window and is available to download on Hugging Face and Kaggle.Google
- Google makes Gemini Nano Banana 2.1 generally available Google released Gemini Nano Banana 2.1 (gemini-nano-banana-2.1) for general availability as an image generation and conversational editing model. It supports 1K, 2K, and 4K resolutions, including panoramic aspect ratios.Google
Oct 5, 2026
- Perplexity launches Decisions API The new Decisions API uses PPLX Decider V1 27B to return probabilities for yes-or-no, choice, or scored answers. Input costs $0.04 per million tokens; output tokens are free, with no per-request fee.Perplexity
- Perplexity cuts Agent API search costs The Agent API's low, medium, and high presets now use Fast Search for web_search, lowering calls from $2.50 to $1.00 per 1,000 invocations. The xhigh preset still uses standard web search.Perplexity
- Perplexity adds image search to Agent API The new image_search tool finds images online and returns image and source-page URLs. Results can be controlled with limits and domain, format, and safe-search filters; successful calls cost $2.50 per 1,000 invocations.Perplexity
- Perplexity adds document input to Agent API The Agent API now accepts PDF, DOC, DOCX, TXT, and RTF files as input_file content. Documents can be sent inline or provided through a public HTTPS file URL.Perplexity
- Amazon Web Services releases Nova 2.5 Sonic Amazon Nova 2.5 Sonic is generally available in Amazon Bedrock across four AWS Regions. The speech-to-speech model has a 256K context window and costs the same as Nova 2 Sonic.Amazon Web Services
Oct 3, 2026
- xAI retires grok-voice-transcribe-1.0 grok-voice-transcribe-1.0 reached end of life on October 2, 2026. Requests to its slug are routed to grok-voice-transcribe-2.0 at the same price.xAI
Oct 2, 2026
- Alibaba Qwen releases Qwen Image 2.1 Pro Qwen Image 2.1 Pro is now available.Alibaba Qwen
- Google retires Gemini 2.5 Flash Image Gemini 2.5 Flash Image is now retired.Google
Oct 1, 2026
- OpenAI retires GPT-5.4 Cyber GPT-5.4 Cyber is now retired.OpenAI
- OpenAI deprecates TTS-1, TTS-1 HD, GPT-5.4 Nano and 2 more TTS-1, TTS-1 HD, GPT-5.4 Nano, GPT-5.3 Codex and GPT-5.1 are now deprecated and shut down on 2027-01-06.OpenAI
- OpenAI deprecates GPT-4o Mini TTS (2025-12-15) and GPT-4o Mini TTS (2025-03-20) GPT-4o Mini TTS (2025-12-15) and GPT-4o Mini TTS (2025-03-20) are now deprecated and shut down on 2027-01-06.OpenAI
Sep 30, 2026
- Runway adds Eleven V4 to Runway Dev Eleven V4 is available on Runway Dev for speech generation, with scripts up to 2,500 characters. It costs 2.2 credits per 1,000 characters through October 12, 2026 PT, then 5 credits, with a 1-credit minimum.Runway
- Voyage AI releases rerank-3 and Rerank 3 Lite Voyage AI announces rerank-3 and Rerank 3 Lite, which retain the Rerank 2.5 series’ 32K-token context length and pricing. The entry says they improve retrieval quality, especially on long documents and code.Voyage AI
- Google retires Gemini Omni Flash Preview Gemini Omni Flash Preview is now retired.Google
- Google announces Gemini 4 Argon Google announced Gemini 4 Argon, with an output token limit of 1 million. It is rolling out to trusted cyber defenders and is not yet available to developers; Google says developer access will come later.Google
- Cohere releases Embed 5 embedding models Cohere released Embed V5.0 Pro and Embed V5.0 Fast, now available through the Embed API, Microsoft Foundry, Amazon SageMaker, and Model Vault. The models accept text, images, and mixed inputs, support over 100 languages, and have a 128k-token context window.Cohere
- Black Forest Labs releases FLUX 3 Image FLUX 3 Image is available through one endpoint for image generation and editing, with up to ten reference images and output up to 4K. Per-image prices range from $0.041 at 768sq to $0.607 at 4K.Black Forest Labs
- Anthropic schedules Claude Sonnet 4.5 retirement Anthropic plans to retire Claude Sonnet 4.5 (claude-sonnet-4-5-20250929) from the Claude API on November 30, 2026. It recommends migrating to Claude Sonnet 5.5.Anthropic
- Alibaba Cloud Model Studio retires DeepSeek-R1 Distill Llama 8B DeepSeek-R1 Distill Llama 8B is no longer available on Alibaba Cloud Model Studio from 2026-09-30.Alibaba Qwen
Sep 29, 2026
- OpenAI releases GPT-6.1 Sol OpenAI released GPT-6.1 Sol (gpt-6.1-sol) for complex coding and professional work. Standard prices for prompts of up to 272K input tokens are $2 per 1M input tokens, $0.10 cached input and $10 output.OpenAI
- OpenAI adds Ultrafast mode for GPT-6 Astra OpenAI added Ultrafast mode for GPT-6 Astra in the Responses API. API customers can use the ultrafast service tier to reduce the time between generated output tokens, subject to rate limits.OpenAI
- OpenAI adds computer use to the Agents API OpenAI added computer use to the Agents API, allowing agents to complete tasks in an OpenAI-hosted browser. Applications handle website access approvals and sign-in.OpenAI
- Mistral AI retires Leanstral 1.5 and GLM 5.2 Leanstral 1.5 retires on September 30, 2026; Z.ai GLM 5.2 retires on October 31, 2026. Mistral AI recommends Z.ai GLM 5.3 as its replacement at the same price.Mistral AI
- MiniMax releases minimax-m3.1-flash-preview 1M-token context window.MiniMax
- Inception makes Mercury Voice generally available Inception made Mercury Voice generally available to enterprise customers. It supports 128K-token contexts, up to 50K output tokens and three reasoning settings. It costs $0.40 per 1M input tokens and $1.50 per 1M output tokens, with 50% off at launch ($0.20 and $0.75).Inception
Sep 28, 2026
- ElevenLabs releases Eleven V4 and Eleven V4 Turbo Eleven V4 and Eleven V4 Turbo are now available. Eleven V4 supports voice cloning in more than 90 languages; Turbo is intended for real-time use and has approximately 100 ms median inference latency.ElevenLabs
- Anthropic releases Claude Sonnet 5.5 Anthropic launched Claude Sonnet 5.5 (claude-sonnet-5-5) on the Claude API, Amazon Bedrock, Claude Platform on AWS, Google Cloud, and Microsoft Foundry. Context window, output limits, and prices are on its model page.Anthropic
Sep 25, 2026
- Perplexity cuts fast preset search price The Agent API fast preset now uses Fast Search. web_search falls from $2.50 to $1.00 per 1,000 invocations and is about 800 ms faster.Perplexity
- Meituan releases LongCat-2.5 Preview 1M-token context window. On LongCat API: $0.30 input, $1.20 output per 1M tokens.Meituan
Sep 24, 2026
- Perplexity adds Fast Search option Search API and Agent API web_search can set search_type to fast for a lower-latency path at $1.00 per 1,000 requests or invocations. Agent API model tokens are billed separately; search_type web is standard search.Perplexity
- Alibaba Qwen releases Decision Model Preview Decision Model Preview is now available.Alibaba Qwen
Sep 23, 2026
- Meta introduces Muse Realtime Avatar On September 23, 2026, Meta introduced Muse Realtime Avatar. It uses Muse Realtime Voice speech tokens and reference media to generate live, synchronized facial, hand, and full-body video.Meta
- Black Forest Labs releases FLUX 3 Action Black Forest Labs released FLUX 3 Action, an open-weight 7B world-action model from its FLUX 3 backbone, with weights available. On RoboLab-120 it uses under half the parameters of the prior best open model and runs up to 3.95x faster.Black Forest Labs
Sep 22, 2026
- OpenAI releases GPT-6 Sol and GPT-6 Luna OpenAI released GPT-6 Sol (gpt-6-sol) and GPT-6 Luna (gpt-6-luna), reasoning models that take text and images and return text through the Responses and Chat Completions APIs. Standard prices per 1M tokens for prompts of up to 272K input tokens: Sol $2 input, $0.20 cached input, $10 output; Luna $0.10, $0.01 and $0.50.OpenAI
- OpenAI improves prompt caching for GPT-6 OpenAI says GPT-6 models hit the prompt cache more often by default, and discounts now apply to shared prefixes reused within 30 minutes. New tools include a caching dashboard, a diagnostics tool that explains cache misses, explicit cache breakpoints, and changing reasoning effort mid-conversation without losing the cache.OpenAI
- Xiaomi releases MiMo-V2.6 Pro UltraSpeed 1M-token context window. On Xiaomi MiMo API: $4.35 input, $8.70 output per 1M tokens.Xiaomi
- Xiaomi releases MiMo-V2.6 Pro 1M-token context window. On Xiaomi MiMo API: $0.435 input, $0.87 output per 1M tokens.Xiaomi
- Xiaomi releases MiMo-V2.6 Flash 1M-token context window. On Xiaomi MiMo API: $0.14 input, $0.28 output per 1M tokens.Xiaomi
- Google releases Gemini 3.8 Flash TTS models Gemini 3.8 Flash TTS (gemini-3.8-flash-tts) and Gemini 3.8 Flash Lite TTS (gemini-3.8-flash-lite-tts) are generally available, with a Gemini API Voices endpoint. Flash-Lite TTS is meant to replace gemini-3.1-flash-tts-preview.Google
- Anthropic releases Claude Opus 5.5 Anthropic released Claude Opus 5.5 (claude-opus-5-5): 1M context, 128k max output, always-on adaptive thinking, at $4/$20 per MTok (Claude Opus 5 is $5/$25). It is on the Claude API, Amazon Bedrock, AWS, Google Cloud, and Microsoft Foundry.Anthropic
- Anthropic previews fast mode for Claude Opus 5.5 Fast mode is available as a research preview for Claude Opus 5.5 on the Claude API.Anthropic
- Anthropic adds mid-conversation tool definitions On the Claude API, the inline-tools-2026-09-15 beta header lets a system message added mid-conversation carry a full tool definition or, with the MCP connector beta header, an MCP toolset, without editing the tools list.Anthropic
Sep 21, 2026
- Alibaba Qwen releases Qwen3.8 Omni Flash Realtime 197K-token context window.Alibaba Qwen
- xAI releases Grok 4.7 500K-token context window. On xAI API: $2.00 input, $6.00 output per 1M tokens.xAI
Sep 20, 2026
- Alibaba Qwen releases Qwen Audio 3.1 Realtime Plus 262K-token context window.Alibaba Qwen
- Qwen releases Qwen Image 2.1 Qwen open-sourced Qwen Image 2.1 on September 20, 2026. The 7B-parameter image model combines text-to-image generation and image editing, with native support for transparent images.Alibaba Qwen
Sep 18, 2026
- xAI releases Grok Voice Transcribe 2.0 Grok Voice Transcribe 2.0 is a speech-to-text model available for batch and streaming transcription. Batch costs $0.10 per hour of audio and streaming costs $0.20 per hour; both include diarization, timestamps, and key terms.xAI
- Google limits access to Gemini 2.5 models Google is limiting Gemini 2.5 API access to users who have actively used those models. They are not deprecated and remain available until further notice. New projects should use 3.5 Flash-Lite or 3.8 Flash.Google
- Qwen releases Qwen3.8-LiveTranslate Qwen announced Qwen3.8-LiveTranslate on September 18, 2026, describing an Interleave architecture for real-time simultaneous interpretation. Its average lagging metric drops from 2.8 seconds to 2.3 seconds.Alibaba Qwen
Sep 17, 2026
- Runway adds Enhance Frame Rate on Dev Runway Dev can now convert video to 24, 25, 30, 48, 50, 60, 120, 23.98, 29.97, or 59.94 fps. Inputs are at most 300 seconds, billed at 1 credit per 2 seconds via the video upscale endpoint with model enhance_frame_rate.Runway
- Alibaba Qwen releases Qwen3.8 Omni Flash 1M-token context window. On Alibaba Cloud Model Studio: $0.15 input, $0.47 output per 1M tokens.Alibaba Qwen
- Alibaba Qwen releases Happyoyster 1.0 Acting Happyoyster 1.0 Acting is now available.Alibaba Qwen
- Google releases Antigravity Agent 09-2026 Google released antigravity-preview-09-2026, replacing antigravity-preview-05-2026. Local tool parameters and file edits changed; remote output-only use needs only the new agent string. The May preview shuts down on October 5, 2026.Google
Sep 16, 2026
- Alibaba Qwen releases Qwen MT Uni Qwen MT Uni is now available.Alibaba Qwen
Sep 15, 2026
- Kling AI releases Virtual Try-On 3.0 Virtual Try-On 3.0 improves face consistency and image quality. It accepts flat-lay, mannequin, and on-model clothing images, and can lock pose, keep the face, and choose whether to keep the original background. V1 and V1.5 parameters map to the 3.0 pipeline.Kuaishou Kling
- TypeSafe releases Jev (preview) 64K-token context window. On TypeSafe API: $0.042 input, $0.00 output per 1M tokens.TypeSafe
- Google releases Gemini 3.8 Live models Gemini 3.8 Live (gemini-3.8-live) and Gemini 3.8 Live Extended Thinking (gemini-3.8-live-extended-thinking) are generally available as audio-to-audio models on the Live API.Google
Sep 14, 2026
- Anthropic adds on-demand compaction to Messages API Anthropic's Messages API now supports on-demand conversation compaction in beta via the compact-2026-09-04 header. A compaction parameter returns a signed summary block you can send later instead of the original messages, while keeping recent turns verbatim.Anthropic
Sep 13, 2026
- Perplexity adds custom MCP connectors A Project can register a remote MCP server once and Perplexity stores its credential. Agent API calls use the connector ID with type connector. API-key or no auth, and Streamable HTTP or SSE, are supported.Perplexity
Sep 11, 2026
- ElevenLabs makes Scribe V2 Medical generally available Scribe V2 Medical is now generally available for batch medical and clinical speech recognition. It is billed at the same rate as Scribe v2.ElevenLabs
Sep 10, 2026
- OpenAI releases Agents API in public beta OpenAI released the Agents API in public beta, with a managed Codex harness for session orchestration, context compaction, and recovery. Sessions can stream progress, use custom tools and MCP servers, and run in OpenAI or customer sandboxes.OpenAI
- OpenAI makes GPT Live 1 generally available GPT Live 1 is generally available for full-duplex voice sessions that continue while a backend model or agent handles reasoning and tools. Voice costs $0.05 per minute, billed per second; model and tool use is charged separately.OpenAI
- DeepSeek releases DeepSeek-V4.1 Flash DeepSeek released DeepSeek-V4.1 Flash, its smallest new-architecture model with native multimodal vision, callable as deepseek-flash. The retired names deepseek-v4-flash and deepseek-v4-flash-vision-exp route to it for now. V4 Pro continues after 14 September 2026 with unchanged billing, and API prices were reduced with the release.DeepSeek
- DeepSeek releases DeepSeek-V4.1 Flash (2026-09-10) DeepSeek-V4.1 Flash (2026-09-10) is now available.DeepSeek
- Black Forest Labs adds 2K and 4K FLUX 3 Video The FLUX 3 Video endpoint now returns qhd (2560×1440) and uhd (3840×2176) clips in one request, and draft_enhance can commit a draft at either size. Added per-second prices are $0.40 and $0.80 for text or image input, and $0.65 and $0.95 for continuation.Black Forest Labs
Sep 8, 2026
- OpenAI releases GPT Image 2.5 Sunburst and Flare OpenAI released GPT Image 2.5 Sunburst and GPT Image 2.5 Flare for generation and editing via the Image API and Responses image tool. Both add xhigh and max quality and use GPT Image 2 token rates.OpenAI
- OpenAI makes GPT-Rosalind generally available OpenAI made GPT-Rosalind (GPT Rosalind Research) generally available for approved internal life sciences research under its trusted-access program. Prices are $5, $0.50 cached, and $25 per 1M input, cached input, and output tokens. Billing starts October 5, 2026.OpenAI
- Inception releases Mercury 2.5 Inception released Mercury 2.5, a quality step up from Mercury 2, with a 260K-token context and 1,107 tokens per second. List price is $0.20 per million input tokens and $0.75 per million output tokens, 80% off at launch.Inception
Sep 7, 2026
- Kling AI launches Video Commerce API Kling AI launched Video Commerce via API and Agent. A character image plus script makes a talking-head video; adding product images produces promotion shots. Options cover audio speed, aspect ratio, resolution, voice, speech rate, and background music.Kuaishou Kling
Sep 4, 2026
- Ant Group releases Ling-3.0 Flash VL On September 4, 2026, Ant Group launched multimodal Ling-3.0-flash-VL. It can be tried in the chat UI and called through OpenAI-compatible and Anthropic-compatible APIs.Ant Group
Sep 3, 2026
- OpenAI releases GPT-6 Astra OpenAI released GPT-6 Astra on the Responses and Chat Completions APIs for reasoning, coding, computer use, research, and documents. It rejects none reasoning effort, custom temperature, top_p, and logprobs. Tool calling requires the Responses API.OpenAI
- OpenAI adds long-running controls for GPT-6 Astra The Responses API now offers async tool calling, mid-turn steering over WebSockets, and mid-conversation reasoning-effort changes for GPT-6 Astra, while preserving the cached prompt prefix.OpenAI
- Google releases Lyria 3.5 music model Lyria 3.5 (lyria-3.5) is generally available for full-length song generation. It accepts text and image inputs and outputs 44.1 kHz stereo audio, with duration and structure controls.Google
Sep 2, 2026
- Alibaba Qwen releases Qwen3.8 Max (2026-09-02) 1M-token context window. On Alibaba Cloud Model Studio: ¥12.00 input, ¥36.00 output per 1M tokens.Alibaba Qwen
- Meta releases Muse Spark 1.3 On September 2, 2026, Meta released Muse Spark 1.3, including a max-reasoning option, on Muse Code and Meta Model API. It targets stronger agentic and coding work and more reliable long instructions.Meta
- Google releases Gemini 3.8 Flash gemini-3.8-flash is generally available. Google describes it as its most intelligent Flash model, aimed at long-horizon software engineering, agents, and enterprise workflows.Google
Sep 1, 2026
- Voyage AI releases Rerank 3 Rerank 3 is now available.Voyage AI
- Voyage AI releases Rerank 3 Lite Rerank 3 Lite is now available.Voyage AI
- Alibaba Qwen releases Qwen3.7 Text Rerank Qwen3.7 Text Rerank is now available.Alibaba Qwen
- Alibaba Qwen releases Qwen3.7 Text Embedding Flash Qwen3.7 Text Embedding Flash is now available.Alibaba Qwen
- Meta releases Muse Voice Transcribe On September 1, 2026, Meta released Muse Voice Transcribe, a real-time audio model with streaming speech recognition, diarization for more than 20 speakers, endpointing, and multilingual code-switching.Meta
- Google adds agentic video understanding to Gemini Agentic video understanding is available for Gemini 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite on the Interactions and GenerateContent APIs. Google says it can use up to 88% fewer tokens on long-form video than static processing.Google
- Anthropic releases Claude Mythos 5.1 1M-token context window.Anthropic
- Anthropic releases Claude Fable 5.1 Anthropic launched Claude Fable 5.1 (claude-fable-5-1), successor to Claude Fable 5. Context is 1M tokens, max output 128k, price $10/$50 per MTok, cache reads $0.25 per MTok, on Claude API, Bedrock, AWS, Google Cloud, and Microsoft Foundry.Anthropic
- Anthropic prices Claude Fable 5.1 cache reads Cache-read price for Claude Fable 5.1 and Claude Mythos 5.1 is $0.25 per million tokens, 0.025 times base input, versus 0.1 times on other models. Cache write prices are unchanged.Anthropic
Aug 31, 2026
- Mistral AI makes OCR 4.1 generally available On August 31, 2026, Mistral AI made OCR 4.1 (mistral-ocr-4-1) generally available.Mistral AI
Aug 28, 2026
- Alibaba Qwen releases Qwen MT Image 2.0 Qwen MT Image 2.0 is now available.Alibaba Qwen
- Tencent releases Hy4 Preview Preview 1.02M-token context window. On Tencent Cloud TokenHub: ¥6.00 input, ¥18.00 output per 1M tokens.Tencent
Aug 27, 2026
- Google releases Gemini Omni Flash Gemini Omni Flash is now available.Google
Aug 26, 2026
- Alibaba Qwen releases Qwen3.8 Flash 1M-token context window. On Alibaba Cloud Model Studio: $0.15 input, $0.47 output per 1M tokens.Alibaba Qwen
- OpenAI deprecates Whisper 1, GPT-4o Transcribe, GPT-4o Mini Transcribe and 1 more Whisper 1, GPT-4o Transcribe, GPT-4o Mini Transcribe and GPT-4o Transcribe Diarize are now deprecated and shut down on 2027-02-26.OpenAI
- Zhipu AI releases GLM-5.3 Flash 1M-token context window. On Z.ai API: $0.15 input, $0.50 output per 1M tokens.Zhipu AI
- Google releases Gemini 3.5 Transcribe Gemini 3.5 Transcribe is now available.Google
- Google releases Gemini 3.5 Transcribe Live Gemini 3.5 Transcribe Live is now available.Google
Aug 20, 2026
- Alibaba Qwen releases Wan3.0 Video Prime Wan3.0 Video Prime is now available.Alibaba Qwen
Aug 18, 2026
- Zhipu AI releases GLM-5.3 1M-token context window. On Z.ai API: $1.40 input, $4.40 output per 1M tokens.Zhipu AI
Aug 17, 2026
- Alibaba Qwen releases Qwen3.8 27B 1M-token context window. On Alibaba Cloud Model Studio: $0.50 input, $3.00 output per 1M tokens.Alibaba Qwen
Aug 13, 2026
- Google releases Gemini 3.7 Flash 1.05M-token context window. On Gemini API: $0.75 input, $3.75 output per 1M tokens.Google
Aug 12, 2026
- xAI releases Grok 4.6 500K-token context window. On xAI API: $2.00 input, $6.00 output per 1M tokens.xAI
Written from verified changes in our dataset or from providers’ official announcements, in our words, and published automatically with source links.