Skip to content
DigIntoAI/Model intelligence
Model intelligence feed

Model intelligence.

Follow the changes. Understand what comes next.

Editor’s pickGoogle
Google/Gemini 4 Neon

Google Launches Gemini 4 Neon in Antigravity With 1M Context, Four Thinking Levels, and Benchmarks vs. Claude Opus 5.5 and GPT-6.2 Sol

Announced on November 19, 2026 at 05:51 Beijing time, Google's Gemini 4 Neon features a 1M-token context window, four thinking levels, and planner-executor pairing with Gemini 4 Hydrogen in Google Antigravity, outperforming Claude Opus 5.5 and GPT-6.2 Sol across most of 18 published benchmarks.

Latest stories349

Newest published
121

Cat Wu Joins OpenAI to Lead Core Products Product Management

Cat Wu (@_catwu) posted on 2026-11-05 08:25:34 Asia/Shanghai that Cat is joining OpenAI to lead product management for Core Products. OpenAI’s Tibo (@thsottiaux) quoted a welcome minutes later and followed with a short “CatGPT” joke. DigIntoAI has no OpenAI HR press release and has not tested any product change—this brief follows the public X posts only.

ChatGPTRead story
122

OpenAI Launches Images 2.6 With GIFs, Reference Images, and GPT-Image-2.6 API Models

OpenAI’s 2026-11-04 Images 2.6: global that day in ChatGPT/Codex on all subscription tiers; ~50% faster. @OpenAI 07:35:14 Asia/Shanghai pins GIFs + reference images. API gpt-image-2.6-flare / gpt-image-2.6-sunburst; same $/1M text $5/$1.25, image in $8/$2, out $30; markup, SVG, complexity. Tibo (2026-11-06): ChatGPT+API >~4B/week, up from >~3B ~8 weeks earlier (self-count >~500M/day). Official / oral self-reports; we haven't hands-on tested.

ChatGPT/Codex/GPT-Image-2.6 Flare/SunburstRead story
123

Paid ChatGPT Message Errors Rise on OpenAI Status—Now Resolved

OpenAI status API opened Elevated error rates across paid ChatGPT plans (id 01M6MBCX427A7113A3T4FJ1N6Y) at 2026-11-05 06:38:12 Asia/Shanghai—impact minor. Some paid-plan users may see errors when sending messages. At 06:49:34, status moved to monitoring after a mitigation. At 08:32:37 Asia/Shanghai (2026-11-05T00:32:37Z), it became resolved; the API says affected services are operating normally again. Overview remains All Systems Operational. DigIntoAI has not tested paid messaging—figures follow the OpenAI status page / status API self-report.

ChatGPTRead story
124

Gemini CLI 0.67.0 Adds Models Command and IDE Sandbox Bridge

Google shipped Gemini CLI 0.67.0 (GitHub v0.67.0, published_at 2026-11-04T20:27:11Z / 2026-11-05 04:27:11 Asia/Shanghai). npm @google/gemini-cli latest=0.67.0 (time 2026-11-04T20:32:45.080Z / 04:32:45 Asia/Shanghai); preview=0.68.0-preview.0 (separate—not merged into 0.67/latest); nightly=0.67.0-nightly.20261104.g89e9d25d6. Headliners: models + --list-models (#29880); IDE companion bridged into container sandboxes over exec stdio (#29858). DigIntoAI has not installed or exercised this build.

Gemini CLI/0.67.0Read story
125

Anthropic Uses Claude to Map the First Prokaryotic Vault Particle Structure

Anthropic’s Science team published Investigating vault particles with Claude (publishedOn=2026-11-04; modified 2026-11-05 03:51:49 Asia/Shanghai): with Claude’s help, the lab reports the first 3D structure of a prokaryotic / phage-related vault particle. Vault function remains unknown. @AnthropicAI (2026-11-05 04:11:28 Asia/Shanghai) links the post and a CDN preprint PDF (~11.3 MB); bioRxiv still pending. Not a new Claude model or pricing change. DigIntoAI has not reproduced the wet lab, re-run cryo-EM, or peer-reviewed the PDF—figures below are self-reported.

Claude/Science/biology research lab/Claude ScienceRead story
126

Gemini 3.8 Flash TTS Voice Remixing Is Live—Tune Timbre With Plain Language

Google says Voice Remixing for Gemini 3.8 Flash TTS is live in AI Studio and the Gemini API: start from 2,000+ or custom voices and fine-tune timbre, pitch, pace, and accent with plain-language prompts. Docs: POST /v1beta/voices with base_voice + relative edits → new stored voice_... for gemini-3.8-flash-tts / gemini-3.8-flash-lite-tts, without overwriting the original. Not a new model release. The 2026-09-23 product blog still shows Coming soon; availability follows posts + docs. DigIntoAI hasn’t tested Studio, the API, listening, or billing.

Gemini 3.8 Flash TTS/Voice Remixing/Google AI Studio/gemini-3.8-flash-tts/gemini-3.8-flash-lite-ttsRead story
127

Kimi K3.1 Pro Open Weights Are on Hugging Face—2.8T / 104B, 1M Context

HF moonshotai/Kimi-K3.1-Pro is public (gated=false): MoE 2.8T/104B, ~1M context, Kimi K3.1 Pro License, ~1.56TB. HF stamp 2026-11-02 12:13:51 Asia/Shanghai. @Kimi_Moonshot released weights 2026-11-05 00:00:35 Asia/Shanghai (links HF). ModelScope Id 1627422 finished SDK batch 2/2 (~1.56TB; 96/96+index)—domestic mirror matches HF; HF remains the release anchor. Tech blog still “by Nov 4” (dateModified 2026-10-24). DigIntoAI has not downloaded or run the weights. Checked 2026-11-05 Asia/Shanghai.

Kimi K3.1 Pro/moonshotai/Kimi-K3.1-Pro/Hugging Face/vLLM/SGLang/TokenSpeedRead story
128

Tencent Hy4 Hits TokenHub and OpenRouter—Vision In, 50% Off Through Year-End

TokenHub listed formal Hy4 (hy4) 2026-11-04 12:13:30, adding image understanding vs preview; 1M / 960k / 64k; list 6 / 18 / 0.3 RMB/M with 50% through 2026-12-31 23:59:59 Beijing time (~3 / 9 estimate). OpenRouter tencent/hy4 2026-11-04 12:20:31 Asia/Shanghai: ~$0.417 / $1.251/M, 50% off, text+image→text. Hy4 preview (hy4-preview) leaves TokenHub/ADP/CloudBase from 2026-12-07 00:00:00 Beijing time; conditional auto-switch to hy4, else call errors.

Hy4/hy4/Hy4 preview/TokenHub/OpenRouter tencent/hy4Read story
129

Claude Text Watermarks Now Cover Mythos 5, Opus 4.7, and Sonnet 4.6—Detection Still Private

On 2026-11-03, Claude Platform release notes state that text from Claude Mythos 5, Claude Opus 4.7, and Claude Sonnet 4.6 carries Anthropic’s text watermark—an expansion of coverage, not a brand-new marking stack. Help Center tables list those models (and others); Amazon Bedrock rollout completes by 2026-11-10; Sonnet 4.6 with non-default temperature/top_p/top_k may omit the mark. Watermark detection remains private preview for eligible organizations. DigIntoAI has not retested detection, sampling edges, or Bedrock region progress.

Claude Mythos 5/Claude Opus 4.7/Claude Sonnet 4.6/Text watermarkRead story
130

OpenAI Plans Next Security Sprint and Raises Bug Bounty Rewards

On 2026-11-03 ~10:29 Asia/Shanghai, OpenAI plans a next internal security sprint with Cathedral Systems and raises Security Bug Bounty rewards. Blogs (November 2, 2026): midpoints for medium/high/critical more than quadrupling on average; ChatGPT/Codex/API critical up to $30,000; exceptional eligible up to $172,000 with bonuses. Make-ups for awaiting awards and past-30-day payouts. DigIntoAI hasn't re-run the bounty math or reproduced vulnerabilities.

Security Bug Bounty/ChatGPT/Codex/API PlatformRead story
131

OpenAI Launches LifeVal Benchmark for Everyday Personal Tasks

On 2026-11-03 ~01:27 Asia/Shanghai (2026-11-02 17:27 UTC), OpenAI announced LifeVal—a benchmark for everyday personal-assistant tasks shaped by feedback from thousands of ChatGPT users. The blog (November 2, 2026) adds a demographically representative survey of 3,995 respondents and says the eval has been used over the past year to track and drive improvement on browser use, computer use, and connectors. Not a new model or pricing drop. We haven't re-tested the blog's example scores, and we haven't seen a full open task set.

LifeVal/ChatGPT(GPT-6 Astra=官博例表提及,非本条上架)Read story
132

Tibo Says Dots Will Expand to GPT-6 Astra and GPT-6.1 Terra

On 2026-11-02 (23:50 Asia/Shanghai / 15:50 UTC), Tibo (@thsottiaux; Codex & ChatGPT @OpenAI) said OpenAI is working on expanding the frontier models that power dots, including GPT-6 Astra, GPT-6.1 Terra, and more later. Terra targets speed, a personality upgrade, and broader dots reach. Follow-up: a dot will pick the best model for the task and may use different models across the work. In progress—not every dot already on Astra/Terra. We haven't tested live dots routing.

ChatGPT dots/GPT-6 Astra/GPT-6.1 TerraRead story