Data Points | The Batch

Engineering lab with multiple screens showing 3D mechanical designs, researchers working with advanced software.

Data Points

Gemini 3 Flash delivers fast, low-cost reasoning: Tencent’s video world model maintains scenes and geometries

GPT’s leaderboard-topping Image 1.5. Qwen’s 30B long-context reasoning model. iRobot’s bankruptcy and planned acquisition. Microsoft’s Voxel 3D generation model.

Small robot with blue eyes points light at open dictionary page, highlighting text.

Data Points

Gemini makes Deep Research available via API: Disney and OpenAI form a Sora licensing deal

Nvidia’s Nemotron 3 family of models, including a 30B standout. Microsoft’s RL-powered Agent Lightning. Google’s new Flash voice models and translation agents. “Slop,” 2025’s word of the year.

Man in field with net catching digital bugs flying from computer screen on table.

Data Points

Anthropic, OpenAI donate open-source projects: GPT-5.2 arrives with new document creation abilities

Mistral’s Devstral 2’s cost-effective coding. GLM-4.6V’s addition of native function calling to vision-language reasoning. Cursor 2.2’s debug and web developer modes. A new executive order preempting local AI regulations in the U.S.

Person uses handheld scanner on historical document, highlighting text.

Data Points

Gemini 3 adds compute-intensive Deep Think: Two famous math problems solved by 10-person AI startup

An executive order that would challenge state AI regulations in the U.S. An 8B parameter open model that shines at coding and math tasks. Publishers’ copyright and trademark lawsuits against Perplexity. Gemini 3 Pro’s ability to visually recognize handwriting from any period.

Students in a classroom building structures with colorful blocks, projector and chalkboard in the background.

Data Points

Mistral 3 update adds four new open models: Amazon Nova 2 family competes with low prices, new agents

Nvidia’s open VLA reasoning model for self-driving cars. Hugging Face’s Claude skills pack that fine-tunes language models. LangChain’s LangSmith no-code, all-chat agent builder. MCP Blockly, a project for students to build their own MCP servers.

Meteorological team in office tracking hurricanes; data displayed on large wall screens for forecasting.

Data Points

DeepSeek’s latest open models rival GPT-5.1 and Gemini 3 Pro: Google’s weather prediction model shined in hurricane season

The new MCP spec’s updates for security and long task support. Runway’s new video generation model’s improved physics. The impact of Character.AI blocking minors from chat. Antigravity, Google’s agentic IDE that rivals Cursor and Windsurf.

Classroom setting with mathematicians working on math equations and examining data graphs.

Data Points

Opus 4.5 drops prices, reclaims coding crown: DeepSeek’s new open-weights math model claims IMO Gold

FLUX.2, a Nano Banana-class open image generator. OpenAI’s third-party web analytics security breach. Genesis Mission, a new U.S. government resource sharing program. Suno’s latest deal with music giant Warner.

Technicians monitor holographic ballet dancers on stage via futuristic screens and robotic controls in a theater setting.

Data Points

Inside Olmo 3, a new family of fully open models: Grok 4.1’s uneasy balance between EQ and sycophancy

Nano Banana Pro, Google’s updated image generator. Anthropic’s latest partnerships with Microsoft and Nvidia. Memo, a home robot trained on real-life human tasks. A new AI play modeled on legendary French playwright Molière’s work.

Japanese computer scientists pointing at a screen

Data Points

Meta model detects and segments video objects: Google Gemini 3 wows on benchmark tests and leaderboards

GPT-5.1-Codex-Max, OpenAI’s improved long-context coding model. Music startup Klay’s reported deal with Universal, Warner, and Sony. DeepSeek R1 Slim, a trim, decensored reasoning model. NTT’s Tsuzumi 2, an efficient model optimized for Japan.

A large screen showing a conceptual map of a sparse neural network, surrounded by a team of computer scientists

Data Points

OpenAI looks inside neural networks: Baidu’s multimodal ERNIE 5.0 arrives, priced to compete

VibeThinker-1.5B, a small but powerful reasoning model. Toymakers’ recall of AI dolls that tell kids how to start fires. Qwen3-Max’s discounts and the latest AI price war. SIMA 2, Google’s self-learning game-playing model.

Blue glass sphere balancing on marble stairs.

Data Points

Generating persistent, editable 3D worlds: Exploring the limits of small synthetic datasets

OpenAI’s GPT-5.1, a more personalizable agentic model. Project FETCH’s use of Claude to program robots. Baidu’s new VL model that tops Gemini Pro and GPT-5. OpenAI liable for copyrighted song lyrics.

Two young adults in a conversation, one African and the other southeast Asian, in traditional dress, wearing microphone headsets

Data Points

Meta AI now recognizes 1600 languages: Amazon and Perplexity spar over browser agents

Kimi K2 Thinking, the new top open model. Deep Research’s expansion to personal documents. Terminal-Bench makers’ new agentic benchmark. TSMC’s slowing revenue growth.