The Inference Report
October 22, 2026
AI Labs — News
OpenAI is moving upstream in enterprise workflows while simultaneously expanding distribution channels. AT&T's legal operations, Google Cloud integration, and ChatGPT Ads reaching 26 new markets signal a shift from consumer-facing novelty to operational infrastructure. The company is bundling safeguards language with distribution agreements, which reads less as safety governance and more as liability management wrapped in policy consistency. Patch the Planet, the security coordination play, is strategic positioning: OpenAI becomes the central hub for vulnerability disclosure and remediation, embedding itself deeper into the dependency chain. The biological data agents work and the advertising expansion are different tactics aimed at the same goal: making OpenAI indispensable to how organizations actually work. None of this requires breakthrough capability. It requires distribution, partnerships, and being present when a company needs to solve a concrete problem.
NVIDIA and AMD are racing to own the infrastructure layer where everything else runs. NVIDIA's announcements cluster around world models, power optimization, and open-weight deployment across finance, retail, energy, and public sector, each one a beachhead in a vertical. The company is explicitly positioning open models as a path for governments and enterprises to maintain control while running on NVIDIA hardware. AMD is responding in parallel, optimizing MoE training and world action models on its MI355X chips. Hugging Face sits in the middle, publishing optimization results and benchmarks that make the hardware differences legible. The real competition isn't about models anymore. It's about who owns the stack that runs them. OpenAI sells access to models. NVIDIA and AMD sell the machines those models run on, and they're winning the margin game by proving that open models deployed on their infrastructure outperform managed services for specific workloads. That's where the money moves next.
Sloane Duvall
AMD
Hugging Face
- Ensemble forecasting with IBM Granite time-series models
- NVIDIA Kumo Relational Outperforms Supervised Machine Learning on Relational Data, No Training Required
- How Specializing the Vocabulary of SLMs Can Cut Inference Cost by 77% in Real Applications
- Profiling TorchNeuron (Part 1): System-level profiling on AWS Trainium.
- Granite Speech 5.0: Transcribe, Translate, and Attribute more Accurately and Faster
NVIDIA
- Leading European Financial Services Organizations Embrace Open Models to Build Proprietary Intelligence
- NVIDIA Cosmos Opens New Tools to Train, Evaluate and Deploy World Models
- How Companies Are Using NVIDIA DSX to Break the Power Bottleneck
- NVIDIA Transforms European Retail, From Digital Discovery to Physical Operations
- Pioneers Across Europe Deploy NVIDIA Earth-2 Open Models to Power Real-World Weather and Climate Prediction
- NVIDIA Open Models Power a New Wave of European Public Sector Projects
OpenAI
- How AT&T is rethinking in-house legal work with OpenAI
- Broad access through consistent standards
- OpenAI models are coming to Google Cloud
- Patch the Planet: global response coordination as agents uncover critical vulnerabilities
- ChatGPT Ads expands to 26 new countries as over 100,000 advertisers have used the platform
- Using thousands of AI agents to uncover new leads in biological data
AI Labs — Models
A curated reference of models from major AI labs, with open/closed weight status, input modalities, and context window size. American labs tend towards closed weights models and Chinese labs tend toward open weights models.
AllenAIClosed Weights
None
Open Weights
- AllenAI: Olmo 3 32B ThinkText66K
AmazonClosed Weights
- Amazon: Nova 2 LiteTextVisionVideoFiles1M
- Amazon: Nova Premier 1.0TextVision1M
Open Weights
None
AnthropicClosed Weights
- Anthropic: Claude Fable 5TextVisionFiles1M
- Anthropic: Claude Fable 5.1TextVisionFiles1M
- Anthropic: Claude Haiku 5.5TextVisionFiles1M
- Anthropic: Claude Opus 4.5FilesVisionText200K
- Anthropic: Claude Opus 4.6TextVisionFiles1M
- Anthropic: Claude Opus 4.7TextVisionFiles1M
- Anthropic: Claude Opus 4.8TextVisionFiles1M
- Anthropic: Claude Opus 5TextVisionFiles1M
- Anthropic: Claude Opus 5.5TextVisionFiles1M
- Anthropic: Claude Sonnet 4.6TextVisionFiles1M
- Anthropic: Claude Sonnet 5TextVisionFiles1M
- Anthropic: Claude Sonnet 5.5TextVisionFiles1M
Open Weights
None
Arcee AIClosed Weights
None
Open Weights
- Arcee AI: Trinity Large ThinkingText262K
Deep CogitoClosed Weights
None
Open Weights
- Deep Cogito: Cogito v2.1 671BText128K
GoogleClosed Weights
- Google: Gemini 3 Flash PreviewTextVisionFilesAudioVideo1M
- Google: Gemini 3.1 Flash LiteTextVisionVideoFilesAudio1M
- Google: Gemini 3.1 Flash Lite PreviewTextVisionVideoFilesAudio1M
- Google: Gemini 3.1 Pro PreviewAudioFilesVisionTextVideo1M
- Google: Gemini 3.1 Pro Preview Custom ToolsTextAudioVisionVideoFiles1M
- Google: Gemini 3.5 FlashTextVisionVideoFilesAudio1M
- Google: Gemini 3.5 Flash LiteTextVisionVideoFilesAudio1M
- Google: Gemini 3.6 FlashTextVisionVideoFilesAudio1M
- Google: Gemini 3.7 FlashTextVisionVideoFilesAudio1M
- Google: Gemini 3.8 FlashTextVisionVideoFilesAudio1M
- Google: Gemini 4 ArgonAudioFilesVisionTextVideo1M
- Google: Lyria 3 Clip PreviewTextVision1M
- Google: Lyria 3 Pro PreviewTextVision1M
- Google: Nano Banana 2 (Gemini 3.1 Flash Image Preview)VisionText66K
- Google: Nano Banana 2 (Gemini 3.1 Flash Image)VisionText131K
- Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)VisionText66K
- Google: Nano Banana 2.1VisionText66K
- Google: Nano Banana Pro (Gemini 3 Pro Image Preview)VisionText66K
- Google: Nano Banana Pro (Gemini 3 Pro Image)VisionText131K
Open Weights
- Google: Gemma 4 26B A4B VisionTextVideo262K
- Google: Gemma 4 31BVisionTextVideo262K
IBMClosed Weights
None
Open Weights
- IBM: Granite 4.2 8BText131K
InceptionClosed Weights
- Inception: Mercury 2Text128K
- Inception: Mercury 2.5Text260K
- Inception: Mercury 3Text1M
Open Weights
None
Meta (Muse)Closed Weights
- Meta: Muse Spark 1.1TextVisionVideoFiles1M
- Meta: Muse Spark 1.2TextVisionVideoFiles1M
- Meta: Muse Spark 1.2 ContributorTextVisionVideoFiles1M
- Meta: Muse Spark 1.3TextVisionVideoFiles1M
- Meta: Muse Spark 1.3 ContributorTextVisionVideoFiles1M
- Meta: Muse WonderTextVisionVideoFilesAudio1M
Open Weights
- Meta: Muse Glimmer 30BTextVision131K
- Meta: Muse Spark OpenTextVisionVideoFilesAudio1M
NVIDIAClosed Weights
None
Open Weights
- NVIDIA: Nemotron 3 Nano 30B A3BText262K
- NVIDIA: Nemotron 3 SuperText262K
- NVIDIA: Nemotron 3 UltraText262K
- NVIDIA: Nemotron 3.5 LightningText262K
- NVIDIA: Nemotron 3.5 Super VLTextVisionVideo1M
- NVIDIA: SwitchyardText1M
OpenAIClosed Weights
- OpenAI: GPT AudioTextAudio128K
- OpenAI: GPT Audio MiniTextAudio128K
- OpenAI: GPT-5.1VisionTextFiles400K
- OpenAI: GPT-5.1-CodexTextVision400K
- OpenAI: GPT-5.1-Codex-MaxTextVision400K
- OpenAI: GPT-5.1-Codex-MiniVisionText400K
- OpenAI: GPT-5.2FilesVisionText400K
- OpenAI: GPT-5.2 ChatFilesVisionText128K
- OpenAI: GPT-5.2 ProVisionTextFiles400K
- OpenAI: GPT-5.2-CodexTextVision400K
- OpenAI: GPT-5.3-CodexTextVisionFiles400K
- OpenAI: GPT-5.4TextVisionFiles1M
- OpenAI: GPT-5.4 Image 2VisionTextFiles272K
- OpenAI: GPT-5.4 MiniFilesVisionText400K
- OpenAI: GPT-5.4 NanoFilesVisionText400K
- OpenAI: GPT-5.4 ProTextVisionFiles1M
- OpenAI: GPT-5.5FilesVisionText1M
- OpenAI: GPT-5.5 ProFilesVisionText1M
- OpenAI: GPT-5.6 LunaFilesVisionText1M
- OpenAI: GPT-5.6 Luna ProFilesVisionText1M
- OpenAI: GPT-5.6 SolFilesVisionText1M
- OpenAI: GPT-5.6 Sol ProFilesVisionText1M
- OpenAI: GPT-5.6 TerraFilesVisionText1M
- OpenAI: GPT-5.6 Terra ProFilesVisionText1M
- OpenAI: GPT-6 AstraFilesVisionText1M
- OpenAI: GPT-6 Astra ProFilesVisionText1M
- OpenAI: GPT-6 LunaFilesVisionText1M
- OpenAI: GPT-6 Luna ProFilesVisionText1M
- OpenAI: GPT-6 SolFilesVisionText1M
- OpenAI: GPT-6 Sol ProFilesVisionText1M
- OpenAI: GPT-6.1 SolFilesVisionText1M
- OpenAI: GPT-6.1 Sol ProFilesVisionText1M
Open Weights
None
PerceptronClosed Weights
- Perceptron: Perceptron Mk1TextVisionVideo33K
- Perceptron: Perceptron Mk1.5TextVisionVideoAudio37K
Open Weights
None
PerplexityClosed Weights
- Perplexity: Sonar Pro SearchTextVision200K
Open Weights
None
PoolsideClosed Weights
None
Open Weights
- Poolside: Laguna S 2.1Text1M
- Poolside: Laguna XS 2.1Text262K
Reka AIClosed Weights
None
Open Weights
- Reka EdgeVisionTextVideo16K
RelaceClosed Weights
None
Open Weights
- Relace: Relace SearchText256K
Thinking MachinesClosed Weights
None
Open Weights
- Thinking Machines: InklingTextVisionAudio524K
- Thinking Machines: Inkling SmallTextVisionAudio524K
xAIClosed Weights
- SpaceXAI: Grok 4.20TextVisionFiles2M
- SpaceXAI: Grok 4.20 Multi-AgentTextVisionFiles2M
- SpaceXAI: Grok 4.3TextVisionFiles1M
- SpaceXAI: Grok 4.5TextVisionFiles500K
- SpaceXAI: Grok 4.6TextVisionFiles500K
- SpaceXAI: Grok 4.7TextVisionFiles500K
- SpaceXAI: Grok Build 0.1TextVisionFiles256K
Open Weights
None
CohereClosed Weights
- Cohere: Command A+TextVision192K
Open Weights
None
Mistral AIClosed Weights
None
Open Weights
- Mistral: Devstral 2 2512TextFiles262K
- Mistral: Ministral 3 14B 2512TextVision262K
- Mistral: Ministral 3 3B 2512TextVision131K
- Mistral: Ministral 3 8B 2512TextVision262K
- Mistral: Mistral Large 3 2512TextVisionFiles262K
- Mistral: Mistral Large 4TextVision1M
- Mistral: Mistral Medium 3.5TextVisionFiles262K
- Mistral: Mistral Small 4TextVision262K
- Mistral: Voxtral Small 24B 2507TextAudioFiles33K
Alibaba (Qwen)Closed Weights
- Qwen: Qwen3.6 FlashTextVisionVideo1M
- Qwen: Qwen3.6 PlusTextVisionVideo1M
- Qwen: Qwen3.7 FlashTextVisionVideo1M
- Qwen: Qwen3.7 MaxText1M
- Qwen: Qwen3.7 PlusTextVision1M
Open Weights
- Qwen: Qwen3 Coder NextText262K
- Qwen: Qwen3.5 397B A17BTextVisionVideo262K
- Qwen: Qwen3.5-122B-A10BTextVisionVideo262K
- Qwen: Qwen3.5-27BTextVisionVideo262K
- Qwen: Qwen3.5-35B-A3BTextVisionVideo262K
- Qwen: Qwen3.5-9BTextVisionVideo262K
- Qwen: Qwen3.6 27BTextVisionVideo262K
- Qwen: Qwen3.6 35B A3BTextVisionVideo262K
- Qwen: Qwen3.8 2.4T A95BText1M
- Qwen: Qwen3.8 27BTextVisionVideo1M
- Qwen: Qwen3.8 FlashTextVisionVideo1M
- Qwen: Qwen3.8 Max (0902)TextVisionVideo1M
- Qwen: Qwen3.8 Max PrimeTextVisionVideo1M
- Qwen: Qwen3.8 Omni FlashTextVisionAudioVideo1M
ByteDanceClosed Weights
- Seed: Seed 1.6VisionTextVideo262K
- Seed: Seed 1.6 FlashVisionTextVideo262K
- Seed: Seed 2.1 TurboTextVisionVideo262K
- Seed: Seed-2.0-CodeTextVisionVideo262K
- Seed: Seed-2.0-LiteTextVisionVideo262K
- Seed: Seed-2.0-MiniTextVisionVideo262K
Open Weights
None
DeepSeekClosed Weights
None
Open Weights
- DeepSeek: DeepSeek V3.2Text164K
- DeepSeek: DeepSeek V4 Flash 0423Text1M
- DeepSeek: DeepSeek V4 Flash 0731Text1M
- DeepSeek: DeepSeek V4 Flash Vision ExpTextVision1M
- DeepSeek: DeepSeek V4 Pro 0423Text1M
- DeepSeek: DeepSeek V4 Pro 0813Text1M
- DeepSeek: DeepSeek V4.1 FlashTextVision1M
- DeepSeek: DeepSeek V4.1 Flash 1017TextVision1M
MiniMaxClosed Weights
- MiniMax: MiniMax M2-herText66K
Open Weights
- MiniMax: MiniMax M2Text205K
- MiniMax: MiniMax M2.1Text205K
- MiniMax: MiniMax M2.5Text205K
- MiniMax: MiniMax M2.7Text205K
- MiniMax: MiniMax M3TextVisionVideo1M
Moonshot AIClosed Weights
None
Open Weights
- MoonshotAI: Kimi K2 ThinkingText262K
- MoonshotAI: Kimi K2.5TextVision262K
- MoonshotAI: Kimi K2.6TextVision262K
- MoonshotAI: Kimi K2.7 CodeTextVision262K
- MoonshotAI: Kimi K3TextVisionVideo1M
XiaomiClosed Weights
None
Open Weights
- Xiaomi: MiMo-V2.5TextAudioVisionVideo1M
- Xiaomi: MiMo-V2.5-ProText1M
- Xiaomi: MiMo-V2.6-FlashTextVisionVideoAudio1M
- Xiaomi: MiMo-V2.6-ProTextVisionVideoAudio1M
- Xiaomi: MiMo-V2.6-Pro-UltraSpeedTextVisionVideoAudio1M
Z.aiClosed Weights
None
Open Weights
- Z.ai: GLM 4.6VVisionTextVideo131K
- Z.ai: GLM 4.7Text205K
- Z.ai: GLM 4.7 FlashText200K
- Z.ai: GLM 5Text205K
- Z.ai: GLM 5 TurboText203K
- Z.ai: GLM 5.1Text205K
- Z.ai: GLM 5.2Text1M
- Z.ai: GLM 5.3Text1M
- Z.ai: GLM 5.3 FlashTextVisionVideo1M
- Z.ai: GLM 5.3 FlashXTextVisionVideo1M
- Z.ai: GLM 5.3 PrimeText1M
- Z.ai: GLM 5V TurboVisionTextVideo203K
UpstageClosed Weights
- Upstage: Solar Mini 4Text524K
- Upstage: Solar Pro 3Text131K
- Upstage: Solar Pro 4Text524K
Open Weights
None