Breaking Shifts & Architecture Diffs
Real-time autonomous intelligence tracking breaking API changes, deprecation warnings, and architectural capability diffs across all frontier models.
OpenAI and Luma AI are forcing aggressive deprecation pipelines, highlighted by the upcoming September 2026 API removal of Sora v2 and the sunsetting of Ray2 in favor of Ray3 and FLUX.2. Simultaneously, frontier velocity is shifting toward native multimodal reasoning and real-time world simulation, driven by Mistral's unified voice-agent SDK, Google's Genie 4, and Decart's real-time Lucy 6 architecture. Baidu and Fish Audio are expanding open and specialized access via the MIT-licensed 3.34B Unlimited-OCR 2 and the 35B multi-speaker Drama 3 MoE model.
AI summaries refresh at most once every 24 hours. Catalog summaries remain available when AI is unavailable or its daily allowance is used.
Claude Fable 5.2 is listed on OpenRouter.
First listed 2026-11-07 as anthropic/claude-fable-5.2.. Platform listing, not a lab announcement. See the source for availability and details.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Lucy 6 is listed on fal.ai.
First listed 2026-11-05 as decart/lucy-2-6/realtime.. Platform listing, not a lab announcement. See the source for availability and details.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
LongCat 2.5 is listed on Hugging Face.
First listed 2026-11-04 as meituan-longcat/LongCat-2.5.. Platform listing, not a lab announcement. See the source for availability and details.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Genie 4: A new generation of world simulation
Official model announcement. See the source for capabilities and availability.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
GPT-6.1 Terra brings fast and capable computer use to everyone.
OpenAI summary: "Bringing fast and capable computer use to everyone.". Official model announcement. See the source for capabilities and availability.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Baidu releases Unlimited-OCR 2 open weights under MIT license on Hugging Face.
3.34B vision-language OCR checkpoint (arXiv:2610.39308) with 147,000+ downloads in its first week.. Handles dense multi-column document parsing, mathematical formulas, and handwriting recognition.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Kimi K3.1 Pro and Highspeed launch across Kimi apps, Code, and API.
Post-trained from K3 with improvements to coding, 3D creation, and agentic workflows, according to Moonshot.. One-million-token context; Highspeed is advertised as 2–3 times faster.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Voxtral 3 brings new speech-to-text and text-to-speech models and a voice-agent SDK.
Voxtral Small Transcribe 3, Mini Transcribe 3 (4B), Mini Transcribe Realtime 3 and Mini TTS 3.. Voice-agent pipeline combines listening, reasoning, tool calls and speech in one stream.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
DeepSeek V4.1-Flash-1017 lands with integrated Search & Fetch APIs and sub-second reasoning.
Drastically reduces agent memory overhead while maintaining top-tier math and coding benchmarks.. Full open weights available under MIT license.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
MiniMax releases MiniMax-M3.1-Flash multimodal open weights on Hugging Face.
220B-parameter vision-language MoE optimized for fast long-context reasoning.. Ships as open safetensors weights on Hugging Face.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Fish Audio drops Drama 3: 35B MoE emotive speech model with 16-character multi-speaker synthesis.
Directable performance via natural language bracket tags ([whispers], [laughs], [anger], [crying]).. Multi-character dialogue handling up to 16 distinct speaker tags in a single request.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Claude Cyber Shield 2.0 deploys real-time continuous supply-chain audit and dependency invariant verification.
Proactively scans open source registries for behavioral anomalies and malicious code injection.. Generates automated, cryptographically verifiable remediation PRs.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Kling 4.0 launches globally with native 4K 10-bit HDR and Omni Reference.
Single-pass video generation expanded up to 30 continuous seconds.. Omni Reference combines up to 15 reference assets and 10 keyframe controls.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
MiniMax launches H3 Max Director mode with native camera path scripting and 3D relighting.
Enables virtual cinematography with pan, zoom, roll, and dolly trajectory controls.. Dynamic relighting tools allow shifting time-of-day lighting post-generation.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Big Sleep 2.0 discovers autonomous zero-days in kernel subsystems and opens automated patch validation.
First AI agent capable of end-to-end vulnerability reproduction, triage, and verifiable patch drafting.. Expanded from single user-space libraries to complex operating system kernel modules.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Anthropic ships Claude Code 1.2 with automated cross-repository refactoring and git verification.
Direct terminal agent runs builds, tests, git commits, and interactive PR drafting.. Built-in sandbox isolates bash execution and verifies tests before finishing.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Claude Haiku 5.5 ships with 1M context window and near-Opus coding speed.
Expands context to 1,000,000 tokens while maintaining sub-second response times.. Adaptive thinking mechanism scales reasoning compute dynamically.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Ling 3.1 Flash expands reasoning performance and instruction adherence.
repo inclusionAI/Ling-3.1-flash (created 2026-10-13, modified 2026-10-15). Source adapter: hf-api
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Gemini 4 Argon opens to developers, enterprises and consumers.
Available via the API in AI Studio and the Gemini platform, and in the Gemini app for Google AI Ultra and Pro subscribers.. Introductory price of $2 per million input tokens and $10 per million output tokens.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Meta introduces Muse Wonder and Muse Wonder Full, frontier models for agents.
Meta calls them its most capable models to date.. Active Memory carries context across conversations; Active Focus keeps long tasks on track.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Skywork releases Matrix-Game 4.0 interactive world model weights under Apache-2.0.
Controllable video-to-video and action-conditioned world generation on Hugging Face.. Apache-2.0 open-weight release from Skywork AI.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Cartesia Sonic 2 sets real-time voice benchmark with sub-90ms full-duplex conversational synthesis.
Engineered specifically for low-latency agent architectures.. Emotion control slider via WebSocket API.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Suno v4.5 introduces stem-level studio mixing, custom lyrical structure controls, and lossless 48kHz audio.
Allows downloading separated stems (vocals, drums, bass, synth) directly from any generated track.. Enhanced genre blending and acoustic instrument warmth.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Tencent open-sources Hunyuan3D 2.0 with production-ready PBR material generation.
Outputs clean topology ready for game engines (Unreal Engine 5, Unity).. Fast two-stage diffusion generates textured mesh in under 15 seconds.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
AlphaFold 3.2 expands de novo molecular drug design with quantitative binding affinity scoring.
Directly generates novel molecular binders targeting previously undruggable viral proteins.. Integrated structural confidence scoring for RNA-protein complex interactions.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Google DeepMind unveils Genie 2, generating endless interactive playable 3D worlds in real time.
Takes keyboard/controller inputs and generates consistent 3D world physics on the fly.. Capable of simulating complex platformers, racing, and first-person perspectives.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Qwen-Image 2.1 Turbo lands on Hugging Face for few-step image synthesis.
7.1B distilled checkpoint accelerating Qwen-Image 2.1 generation.. Accompanied by Prompt-Editing (PE) text-to-image and image-to-image variants.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Moshi 1.5 drops latency to 110ms with multilingual conversational turn-taking and emotional mirroring.
Native speech-to-speech dialogue with instantaneous human interruption handling.. Open weights and Mimi audio codec updated with enhanced voice timbre clarity.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Oasis 1.5 reaches 60 FPS real-time neural simulation with multi-agent persistent memory.
Fluid gameplay with zero game engine code; all rendering and world mechanics emerge from weights.. Persistent terrain changes survive long-horizon exploration sessions.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
FLUX.2 Turbo drops latency to 8 steps with complete typography prompt adherence.
Ultra-fast distilled weights for real-time graphics pipelines.. Renders multi-line typography with zero spelling distortion.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Mistral Large 4: a 1-trillion-parameter natively multimodal MoE.
52 billion active parameters.. Public preview API live in Mistral Studio, served from European datacenters.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Fish Speech 2.0 introduces sub-100ms streaming TTS.
Zero-shot voice cloning from 3 seconds of audio.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Midjourney V8.5 introduces direct web-canvas camera re-framing and 4K photorealism.
Interactive 3D bounding box controls for multi-character scene consistency.. Significant improvement in complex hand anatomy and legible signage text.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Stable Audio 2.5 expands diffusion style-transfer to full 5-minute continuous audio arrangements.
Multi-track stems export with discrete drum and bass separation.. Trained exclusively on licensed commercial music catalogs.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Ideogram 4.5 is listed on fal.ai.
First listed 2026-09-30 as ideogram/v4.5/edit, ideogram/v4.5.. Platform listing, not a lab announcement. See the source for availability and details.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Gemini 4 Argon announced, rolling out first to trusted cyber defenders.
Early access runs through Google’s Fairwind Program while safeguards are tested before broad availability.
No breaking indicators found in the recorded notes. Compatibility has not been tested; check the provider release notes.
Claude Opus 5.5 takes #1 on frontier reasoning and autonomous software engineering.
Sets state-of-the-art across SWE-bench Verified and competitive coding benchmarks.. Native multi-agent orchestration for continuous 24-hour development runs.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
GPT-6.1 Sol offers near-Astra intelligence at one-fifth of Astra’s API token prices.
OpenAI summary: near-Astra intelligence for coding, computer use and professional work.. Priced at one-fifth of Astra’s standard API input and output token prices.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Eleven v4, ElevenLabs’ most emotive model, launches with a new architecture.
Adds a Turbo model for low-latency use cases.. ElevenLabs cites more authentic, consistent voice cloning across languages.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Claude Sonnet 5.5 is listed on OpenRouter and models.dev.
First listed 2026-09-28 as anthropic/claude-sonnet-5.5, anthropic/claude-sonnet-5.5:batch.. Platform listing, not a lab announcement. See the source for availability and details.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
GPT-6 Sol and Luna bring frontier intelligence to everyday work.
Two models with different balances of capability and cost.. Official model announcement. See the source for capabilities and availability.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Grok 4.7 is listed on OpenRouter and models.dev.
First listed 2026-09-21 as x-ai/grok-4.7.. Platform listing, not a lab announcement. See the source for availability and details.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Veo 3.2 integrates Google DeepMind spatial audio and 4K visual resolution.
Cinematic lighting controls and temporal consistency across scene transitions.. Deep integration with YouTube Shorts and Gemini creator workflows.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
ElevenMusic 2.5 is listed on fal.ai.
First listed 2026-09-14 as elevenlabs/music/v2.5.. Platform listing, not a lab announcement. See the source for availability and details.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Alibaba releases Qwen-Image 2.1 open weights on Hugging Face.
7.1B-parameter text-to-image foundation model with over 220,000 Hub downloads.. Supports bilingual typography rendering and structured prompt adherence.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Fable 5.1 expands multi-agent alignment and complex scientific theorem proving.
Pinnacle intelligence tier exceeding Opus in sustained mathematical rigor.. Strict safety evaluations and verified cyber-safeguards.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Qwen3.8 Max 0902 checkpoint introduces refined multi-turn tool-use.
openrouter lists qwen/qwen3.8-max-0902 ("Qwen: Qwen3.8 Max (0902)") from 2026-09-03. Source adapter: openrouter
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
GPT-6 Astra debuts as OpenAI’s most intelligent and aligned model yet.
OpenAI cites state-of-the-art capabilities across computer use, coding, cybersecurity and science.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Gemini 3.8 Flash and 3.8 Flash Cyber arrive three weeks after 3.7.
Google calls 3.8 its best reasoning and coding model yet, at the speed and cost of 3.7.. 3.8 Flash Cyber is a cybersecurity variant.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Muse Spark 1.3 is listed on OpenRouter and models.dev.
First listed 2026-09-02 as meta/muse-spark-1.3.. Platform listing, not a lab announcement. See the source for availability and details.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Z.ai releases GLM-5.3 and GLM-5.3-Flash open-weight checkpoints on Hugging Face.
753B-parameter MoE flagship alongside a high-throughput Flash variant.. Over 6.3 million downloads across the GLM-5.3 release repositories on Hugging Face.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
DeepSeek R1 matches frontier reasoning with open weights and emergent self-verification chains.
Trained via large-scale reinforcement learning without cold-start SFT.. Distilled dense variants run on single consumer workstations.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Gemini 3.7 Flash, Google’s most intelligent workhorse model yet for coding and agents.
Official model announcement. See the source for capabilities and availability.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Grok 4.6 is listed on OpenRouter and models.dev.
First listed 2026-08-12 as x-ai/grok-4.6.. Platform listing, not a lab announcement. See the source for availability and details.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
LTX 2.5 is listed on fal.ai.
First listed 2026-08-11 as lightricks/ltx-2.5/image-to-video/pro, lightricks/ltx-2.5/image-to-video/fast, lightricks/ltx-2.5/text-to-video/pro.. Platform listing, not a lab announcement. See the source for availability and details.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
MAI-Image-2.6 launches at No. 2 on Arena’s text-to-image leaderboard.
Microsoft reports +79 Elo over MAI-Image-2.5 overall and +91 Elo on text rendering.. MAI-Image-2.6 and 2.6-Flash entered Public Preview on Microsoft Foundry (post updated 4 September 2026).
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Seedream 5.0 Pro adds grounded, layer-separated editing.
Edits a single region while holding the rest of the frame.. Layer separation for compositing.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Tripo3D image-to-3D object generation.
Single image in, 3D model out.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Muse Spark 1.2 ships with Muse Code, a terminal coding agent.
Terminal-Bench 2.1 up to 82.9% from 76.2%; DeepSWE v1.1 to 59.3% from 53.0%.. Trained on whole software engineering workflows rather than completion.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Qwen3.8-Max, 2.4T parameters and the first Max tier to open.
2.4 trillion parameters, 95B active. Text, image and video, 1M context.. First Max-class Qwen to be open-sourced — a return to the open strategy after several proprietary flagships.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
DeepSeek V4 Flash delivers high-throughput reasoning at record-low inference cost.
Flagship speed tier for production agent deployments.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
MiniMax H3 text-to-video lands with high frame consistency and fluid physics.
First generation H3 model release.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Hi3D generates 3D models from a single image.
Direct competitor to Tripo in the image-to-3D slot.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Full K3 weights published on Hugging Face, on the date Moonshot committed to.
Moonshot calls K3 the world’s first open 3T-class model.. Released under Moonshot’s own kimi-k3 licence.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Qwen3.7-Flash lands between the 3.5 train and 3.8-Max.
Flash tier; announcement date not yet confirmed against Alibaba.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Opus 5 lands below Fable 5, at half the price.
New Opus-tier flagship above Opus 4.8.. Anthropic positions it near Fable 5's intelligence at $5/Mtok input against Fable 5's $10.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Ling-3.0-flash released with open weights.
Throughput-oriented MoE.. Part of the July open-weight wave.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
MAI-Image-2.5-Pro introduced alongside MAI-Voice-2-Flash.
Microsoft highlights accurate text rendering and natural-language edits.. Also listed on fal.ai as microsoft/mai-image-2.5-pro.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Three Qwen releases inside 72 hours.
Rapid-fire release train across sizes.. Open weights on all three.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Laguna S 2.1 released into the July model wave.
Incremental gains on repo-scale editing.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
GPT-5.6 arrives as a lineup: Sol, Terra and Luna.
Three named variants released together instead of a single flagship.. Splits reasoning depth, speed and cost across the tiers.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Sonilo V1.1 generates licensed music from one prompt.
Control over style, mood, instrumentation and exact duration.. Licensing is the differentiator, not fidelity.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Gemini 3.6 Flash ships as a three-model trio.
Three Flash variants released simultaneously.. Aimed at high-volume, latency-sensitive workloads.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
LongCat 2.0, a 1.6T-parameter sparse MoE from Meituan.
48B active parameters.. Positioned for coding and repository tasks.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Grok 4.5 pushes coding and knowledge work at lower token use.
Headline claim is efficiency, not raw capability.. Fewer tokens consumed per completed task.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Muse Spark 1.1 leans hard into computer use.
1,000,000-token context window.. Positioned on agent work rather than chat.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Kimi K3 launches at 2.8T parameters with native vision and a 1M-token context.
Built on Kimi Delta Attention and Attention Residuals, per Moonshot.. Available on Kimi.com, Kimi Work, Kimi Code and the Kimi API at launch.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
LTX-2.3 ships text-, image- and audio-to-video.
Pro and Fast variants.. Generates synchronised audio with video.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
FLUX 3 enters early access — video, audio and robot actions from one model.
First BFL model to generate video, audio and robot action-prediction from a single set of weights.. FLUX 3 Video does up to 20-second clips with native audio.
No breaking indicators found in the recorded notes. Compatibility has not been tested; check the provider release notes.
Lucy 2.5 edits webcam and streamed video live.
Restyle, background swap and object replacement at interactive rates.. Runs over WebRTC rather than as a batch job.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Reve 2.1 edits from text prompts with layout intelligence.
Prompt adherence and accurate text rendering.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
KAT-Coder V2.5 lands in Pro and Air sizes.
Both variants shipped the same day.. Scoped at whole issues and workflows.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Seedream 5.0 Lite editing endpoint.
Fast variant, multi-input editing.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Hy3 ships with configurable reasoning.
21B active of 295B total.. 192 experts with top-8 routing.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Ideogram V4.0 sharpens text rendering and fine detail.
Accurate typography remains the selling point.. Posters and logos as the target use.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Sonnet 5 arrives under the Opus 5 flagship.
Tracked separately from Opus — different cadence, different tier.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Nano Banana 2 Lite, also listed as Gemini 3.1 Flash Lite Image.
Same model under two ids — the case the alias map exists for.. Lite tier below Nano Banana 2.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Baidu open-sources Unlimited-OCR with 3.34B MIT-licensed weights on Hugging Face.
Over 4.2 million Hub downloads and 4,900+ likes since release (arXiv:2606.23050).. End-to-end image-text-to-text document recognition in safetensors format.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Seedance 2.5 extends clip length and shot consistency.
Now the strongest long-form image-to-video option.. Better identity retention across cuts.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
GLM-5.2 ships 753B MIT-licensed open weights on Hugging Face.
MIT-licensed open-weight foundation model for reasoning and agentic coding.. Published in BF16 and FP8 weight configurations.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Fable 5 ships as Anthropics premier Mythos-class intelligence engine.
Designed specifically for multi-day uninterrupted agentic task planning.. Introduced prior to the Opus 5 release window.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Ray3.2 adds frame-by-frame direction and a public API.
Clips up to 20 seconds at 1080p.. Up to 16 keyframes per clip to shape motion and pacing.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
TripoSplat converts one image into 3D Gaussians.
Open-source, from TripoAI / VAST AI Research.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
MAI-Image-2.5 and 2.5-Flash launch; 2.5 ranks No. 2 on Arena’s image-edit leaderboard.
Generation and precise, localized editing in one model.. 2.5-Flash targets fast, scalable production workloads.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
MiniMax-M3 ships 427B multimodal MoE weights on Hugging Face.
427B-parameter native image-text-to-text architecture.. Released with BF16 and MXFP8 weight checkpoints.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
PaddleOCR-VL 1.6 upgrades compact 0.96B multilingual document and table parsing on Hugging Face.
Apache-2.0 open weights supporting vLLM and Transformers inference for text, tables, formulas, and charts.. 87,000+ Hub downloads with ultra-low sub-1B parameter footprint.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
First Muse Spark release.
Agent-first framing from launch.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Recraft V4.1 brings quieter photorealism and new illustration styles.
Adds a Utility variant for when simplicity matters more than creative surprise.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
V8.1 is the fastest Midjourney yet — 4–5× on standard jobs.
HD 2K output added.. Better prompt adherence and small-detail retention.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Qwen 3 generation opens with hybrid reasoning.
Thinking and non-thinking modes in one checkpoint.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Hy3 preview, ahead of the full release.
High-efficiency MoE for agentic workflows.
No breaking indicators found in the recorded notes. Compatibility has not been tested; check the provider release notes.
V3 with style training from reference images.
Fine-tuning on a small reference set.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Kimi K2.6 open-sources long-horizon coding and agent swarms.
Available via Kimi.com, the Kimi App, the API and Kimi Code.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
MAI-Image-2-Efficient: flagship quality at nearly 41% lower cost.
Microsoft reports 22% faster generation.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Incremental bump to the 4-series flagship.
Coding and tool-use gains over 4.5.. Same pricing as 4.5.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
MiniMax-M2.7 open-weight 228B MoE lands on Hugging Face.
228B-parameter text reasoning checkpoint preceding the M3 generation.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
ElevenMusic launches on iOS, unifying voice, music and SFX.
One app across voice, music and sound effects.. Positioned against Suno on licensing rather than quality.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Matrix-Game 3.0 open-sources image-and-text-conditioned playable world generation.
Apache-2.0 open weights for interactive environment synthesis.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
v5.5 adds custom fine-tuning and verified voice cloning.
Better vocal expressiveness over v5.. Custom model fine-tuning for users.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Sora 2 deprecated; API removal set for 24 September 2026.
- •API Lifecycle Alert: Removed from the OpenAI API, with no recommended replacement listed. (Target: 2026-09-24)
- •Sora 2 deprecated; API removal set for 24 September 2026.
OpenAI notified developers on 24 March 2026.. The Videos API and every sora-2 and sora-2-pro snapshot leave the API on 24 September 2026.
Potential migration issue: Review source evidence and legacy dependencies on Sora v2 and verify replacement endpoints.
MAI-Image-2 targets natural light, accurate skin tones and lived-in scenes.
Official model announcement. See the source for capabilities and availability.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Gemini 3 generation reaches the Flash tier.
1M-token context standard across the tier.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Laguna S 2 targets whole-repository reasoning.
Trained on execution feedback rather than static code alone.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Nano Banana 2, Google's image generation and editing model.
Generation and editing in one endpoint.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Kling 3.0 adds multilingual lip sync and cinematic motion controls.
Lip sync across multiple languages in one pass.. Holds four separate entries in the video top ten.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
5.2 tightens routing between fast and thinking modes.
Less visible mode-switching.. Lower latency on short turns.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Recraft V4, a ground-up rebuild focused on design taste.
Targets visual taste, prompt accuracy and output quality for production design work.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Seedance 2.0 takes the #1 slot on Artificial Analysis.
Displaces Veo and Kling at the top of the video board.. Large jump in motion coherence.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
PaddleOCR-VL 1.5 ships Apache-2.0 vision-language OCR checkpoints on Hugging Face.
Sub-billion-parameter document layout and OCR model from Baidu's PaddlePaddle team.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Kimi K2.5 turns K2 into a native multimodal model.
Continued pretraining on about 15T mixed visual and text tokens, per Moonshot.. Can direct an agent swarm of up to 100 sub-agents.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Ray3.14: native 1080p, 3x cheaper and 4x faster.
Luma positions it as removing the quality, speed and cost tradeoff of Ray3.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
V8 resets the base model.
Large jump in prompt following.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
FLUX klein targets local and enterprise inference.
40% less memory on NVIDIA RTX cards.. Roughly 40% faster.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
v2 improves stem separation and instrumental control.
Preferred by musicians over Suno on instrumental work.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Runway Gen-4.5 tops the Artificial Analysis text-to-video leaderboard.
Runway cites 1,247 Elo on the leaderboard as of 30 November 2025.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
FLUX.2 replaces FLUX.1 as the flagship.
- •FLUX.2 replaces FLUX.1 as the flagship.
Leads open-weight image quality.. Strong on photoreal and commercial work.
Potential migration issue: Review source evidence and legacy dependencies on FLUX and verify replacement endpoints.
Opus 4.5 lands with a large price cut.
Roughly a third of 4.1 pricing.. Effort control added to the API.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Veo 3.1 brings true 4K at up to 60fps.
3840×2160 output.. Ambient sound, dialogue and effects generated with the video.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
MAI-Image-1 debuts in the top 10 on LMArena.
Later added to Bing Image Creator.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Sora 2 launches with synced audio and a social app.
Best-in-class physical motion.. Shipped with a consumer feed app.
Potential migration issue: Review source evidence and legacy dependencies on Sora and verify replacement endpoints.
2.5 Turbo cuts cost while improving motion.
Faster generation at lower price per second.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
v5 sets the current quality benchmark at 1,293 ELO.
Ahead of every competitor on fidelity, structure and vocal realism.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Luma launches Ray3.
- •New generation of the Ray video model; Ray2 and Dream Machine are now deprecated.
New generation of the Ray video model; Ray2 and Dream Machine are now deprecated.
Potential migration issue: Review source evidence and legacy dependencies on Ray and verify replacement endpoints.
GPT-5 replaces the 4-series across ChatGPT.
- •GPT-5 replaces the 4-series across ChatGPT.
Unified router over reasoning and non-reasoning paths.
Potential migration issue: Review source evidence and legacy dependencies on GPT-5.6 and verify replacement endpoints.
Opus 4.1 refines agentic search and coding.
Drop-in replacement for Opus 4.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
ElevenLabs enters music generation.
Built on licensed training data from the start.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Wan 2.2 releases open weights and inference code.
Text-to-video and image-to-video A14B models plus a 5B text-image-to-video model.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
K2 established the open MoE line.
1T-parameter MoE, agentic tool use.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Grok 4 released with a heavy multi-agent tier.
Grok 4 Heavy runs multiple agents in parallel.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
First public Seedance release.
Multi-shot generation from a single prompt.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
v3 adds audio tags for directed emotional delivery.
Inline tags control emotion and pacing.. 70+ languages.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Veo 3 introduces synchronised native audio.
First major model to generate video and matching audio together.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
v4.5 leads on lyric integration.
Best-in-class lyric-to-melody alignment at release.
Standard enhancement: Test updated capability gains and cost efficiencies against current production benchmarks.
Kling 2.0 rebuilds the motion model.
Large improvement in physical plausibility.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
V7 introduces draft mode and personalisation by default.
Personalisation on by default for the first time.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Runway Gen-4 keeps characters, locations and objects consistent across scenes.
Official model announcement. See the source for capabilities and availability.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Alibaba open-sources Wan 2.1 matching commercial video generators with consumer GPU support.
Supports both T2V and I2V at native 1080P resolution.. 1.3B model runs on consumer RTX 4090 GPUs.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Anthropic deploys specialized Claude cybersecurity defense suite for automated patch verification.
Provides deep binary and source-level security audits with verifiable proof-of-fix.. Rigorous safety evaluation preventing dual-use weaponization.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Big Sleep discovers first-ever in-the-wild zero-day in widely used open source software (SQLite).
Evolved from Project Naptime into a specialized LLM-assisted vulnerability discovery agent.. Discovered stack buffer underflow in SQLite before it reached any production release.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Decart and Etched release Oasis, a real-time neural game engine running Minecraft without game code.
Takes keyboard and mouse inputs and generates the next video frame autoregressively.. Simulates physics, inventory, biome rendering, and ray-traced lighting purely from weights.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Recraft V3 sets a new image-generation standard, led by text rendering.
Recraft cites first place on the Artificial Analysis text-to-image benchmark.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Mistral Large 2 delivers 123B parameter frontier reasoning with 128k context.
State-of-the-art multilingual reasoning rivaling GPT-4o and Llama 3.1 405B.. Engineered to drastically minimize hallucination and sycophancy.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Kyutai open-sources Moshi, achieving conversational speech-to-speech without STT/TTS cascade.
First native audio-to-audio foundation model with realistic emotional prosody and backchanneling.. Full duplex: can listen and speak simultaneously without audio cutoffs.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
AlphaFold 3 predicts the structure and interactions of all of lifes molecules.
Goes beyond proteins to model DNA, RNA, ligands, antibodies, and chemical modifications.. Diffusion-based architecture replacing previous Evoformer.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.
Stable Audio 2.0 introduces audio-to-audio style transfer and 3-minute stereo generation.
Generates full-length structured tracks (intro, development, outro) from text or audio prompts.. Trained on AudioSparx licensed catalog for copyright integrity.
High architectural impact: Conduct staging evaluation and verify prompt/token schemas before production cutover.